(L&N) Q&A Thread - Quick questions go here

[Nel Fie]

I'd also suggest that widespread allomorphic alternations would help build a good case for two allophones being the same phoneme.

[...]

Provided that the rules for determining the surface allophone are reasonably simple (like 'only a full vowel when not adjacent to another vowel, otherwise a glide') it wouldn't take much to make such an analysis look simpler than the alternative.

Very good and useful information - thank you for taking the time to write it all out!

[regenbogen9]

What's the difference between an Abugida and an Alpha-Syllabary?

[WeepingElf]

What's the difference between an Abugida and an Alpha-Syllabary?

I think it depends on whom you ask ;) It seems to me as if both mean more or less the same thing; at least, I consider them synonyms.

[Salmoneus]

According to wikipedia:

- in an abugida, the consonant symbols must have an unwritten 'inherent vowel'; an explicit vowel diacritic is used when the actual vowel differs from the inherent vowel

- in an alphasyllabary, there are symbols for both consonants and vowels (whether or not an inherent vowel exists), and the vowels are written in a way associated with the consonant but not always linearly associated with it in the actual order of speech.

The abugida concept is therefore about the idea of consonant symbols containing vowels, and needing some markers if the contained vowel is 'wrong'. Whereas the alphasyllabary concept is about vowel and consonant symbols forming blocks representing whole syllables, which can be divided into consonant and vowel parts (unlike a syllabary) but without the vowel parts regularly following the consonant parts (so unlike an alphabet).

A system can be BOTH alphasyllabic and abugidic, if it has non-linear vowels AND an unwritten inherent vowel; a system can be an abugida but not an alphasyllabary if is has unwritten inherent vowels but all other vowel markings are linear; a system can be an alphasyllabary but not an abugida if it has non-linear vowels but NO inherent vowel; and a system can be neither if there is no inherent vowel, and either vowels are unmarked (abjad), or are marked linearly (alphabet), or are combined with the consonants in a non-decomposable way.

"Alphasyllabary" forms a more conceptually coherent alternative option with alphabets, syllabaries, and abjads, since those four essentially completely cover non-logographic systems. "Abugida" isn't really part of that set of concepts, since instead it's a practical observation about a common writing technique in practice (usually but not exclusively found in alphasyllabaries).

[Keenir]

According to wikipedia:

- in an abugida, the consonant symbols must have an unwritten 'inherent vowel'; an explicit vowel diacritic is used when the actual vowel differs from the inherent vowel

- in an alphasyllabary, there are symbols for both consonants and vowels (whether or not an inherent vowel exists), and the vowels are written in a way associated with the consonant but not always linearly associated with it in the actual order of speech

I was under the impression that an Alphasyllabary was subgroup/subtype of Syllabary. Is the difference between an Abugida and a Syllabary the same as with an Alphasyllabary?

[Arayaz]

According to wikipedia:

- in an abugida, the consonant symbols must have an unwritten 'inherent vowel'; an explicit vowel diacritic is used when the actual vowel differs from the inherent vowel

- in an alphasyllabary, there are symbols for both consonants and vowels (whether or not an inherent vowel exists), and the vowels are written in a way associated with the consonant but not always linearly associated with it in the actual order of speech

I was under the impression that an Alphasyllabary was subgroup/subtype of Syllabary. Is the difference between an Abugida and a Syllabary the same as with an Alphasyllabary?

In a syllabary, all the glyphs stand for unique syllables and have no connection with one another.
In an alphasyllabary, all the glyphs stand for consonants and have diacritics indicating vowels.
In an abugidea, all the glyphs stand for a consonant plus a particular vowel and have diacritics indicating other vowels.

At least, that's my understanding of it.

[Salmoneus]

Well, let's get conceptual!

The prototypical language is composed of syllables that are, prototypically, CV in structure. Two parts: C, and V.

All non-logographic writing (and arguably even most logographies in practice) can be seen as a sequence of encoded syllables. These syllables are encoded linearly, in a way that iconically represents the sequence of utterance of those syllables: a line of syllable-symbols, from one side (the first) to the other (the last). All nice and simple.

The question is: what relation do those syllable-symbols have to one another?

Easiest option: none. Each syllable has a distinct and unrelated symbol. Let's call this Option A.

Alternatively, there may be a systematic relationship between certain sets of symbols.

Which sets? There are many hypothetical options (POA, MOA, tone, etc).

The most common form of organisation, though, is to link syllables by their onset consonants. Each syllable beginning with a certain consonant has a specific visual element or grapheme indicating its membership of the set of syllables with that onset.

Let's call all other systems of organisation Option B.

If syllable-symbols contain some element to indicate the onset, what about the nucleus (i.e the vowel)?

One option is to not indicate the vowel at all. Option C.

Another is to distinguish different syllables with the same onset but different nuclei in a non-systematic way: Option D.

Or, there could be some systematic way in which syllables with the same nucleus are ALSO linked.

So how? Two (sets of) options.

Either the element indicating the nucleus can be treated in the same way as that of the consonant - that is, as a discrete symbol placed in the ordinary linear flow of symbols, coequal with the consonants, Option E - or it can be different, and in some sense subordinate to the consonant, in that the placement of the vowel symbol in the flow of time is not independent, but is dependent upon the onset consonant, and not itself part of the sequence of linearised symbols: Option F.

Option A is a syllabary.

Option B is... probably also a syllabary.

Option C is an abjad.

Option D... to my knowledge doesn't exist?

Option E is an alphabet.

Option F is an alphasyllabary.

In the alphasyllabary, the onset-nucleus pair could be imagined as a single composite symbol representing the entire syllable (like a syllabary), OR as a pair of symbols reprsenting onset and nucleus (like an alphabet), only not consistently positioned in chronological order.

[there's also the situation where the nucleus symbol is not a graphical symbol but a conceptual one, like the rotation of the consonant, which is I think for convenience's sake considered an alphasyllabary as well]

And an abugida is a system in which some nuclei are indicated but others are not, because a default nucleus is assumed in the absence of overt marking. An abugida in this sense can be either an alphabet or an alphasyllabary, although in practice usually the latter.
Does that explain everything? I wouldn't get too hung up on it, it's just one guy's terminology.

[WeepingElf]

Thank you. So I was mistaken in assuming that they are the same thing.

[Salmoneus]

Yes and no, I think. The terms as originally defined are theoretically overlapping, not identical in meaning. However, I think they are widely treated as synonyms. The wikipedia article treats them as synonymous, except for a couple of paragraphs explaining how technically, as originally used, the terms weren't actually synonymous, and a few writing systems are in strict terms one but not the other.

I think the distinction is useful to bear in mind for a conlanger considering all the possibilities. But for most purposes I think they can be casually used as synonyms or near-synonyms? [the great majority of actual writing systems that are one are also the other]

[WeepingElf]

So an abugida is a specific kind of alphasyllabary, I guess?

[Salmoneus]

So an abugida is a specific kind of alphasyllabary, I guess?

No, an alphabet could also be an abugida. Unless you define it as not an alphabet because it's an abugida. But certainly non-alphasyllabaries can be abugidas, as I've explained.

For instance:

Lorm£ Xip£sum£ dolor£ sit£ Xamt£, con£sc£ttur£ Xadipis£cin£g£ Xlit£, sd£ do XXiXus£mod£ tm£por£ Xin£cididun£t£ Xut labor Xt dolor mag£na XaliquXa.

This text would be an abugida, because it consists of consonants (including a dummy for null onsets) with an inherent vowel (e), with vowels only marked when they deviate from the inherent vowel (including in this case a dummy vowel for codas). However, it would not be an alphasyllabary, because the consonant-vowel symbol blocks are constructed linearly, with vowels not dependent upon consonants or in any way subordinated to them organisationally. Instead, this is an alphabet with explicit coda marking and elision of one vowel, if you still call that an alphabet.

[Keenir]

Well, let's get conceptual!

I need to be sure I understand...

Another is to distinguish different syllables with the same onset but different nuclei in a non-systematic way: Option D.

Or, there could be some systematic way in which syllables with the same nucleus are ALSO linked.

Okay, when I hear "systematic", I think of those charts that have the vowels running in one direction, and the consonants running in the other direction...

__ | -e | -a |
b- |
t- |

...like that.

Either the element indicating the nucleus can be treated in the same way as that of the consonant - that is, as a discrete symbol placed in the ordinary linear flow of symbols, coequal with the consonants, Option E - or it can be different, and in some sense subordinate to the consonant, in that the placement of the vowel symbol in the flow of time is not independent, but is dependent upon the onset consonant, and not itself part of the sequence of linearised symbols: Option F.

Option E is an alphabet.

Option F is an alphasyllabary.

The flow of time? Thats very poetic, kudos, but not very clear to my ears, sorry.

[Salmoneus]

Systematic means having a coherent system.

So, for instance, if you have the syllable-symbol BA and add a mark to make it BAN, and you have a syllable-symbol PA and add the same mark to make it PAN, that's a systematic relationship between symbols.


I wasn't trying to be poetic.


Take a word like "yataghan". We can divide this into symbols representing sounds: Y - A - T - A - GH - A - N. Notice how, if you read those symbols from left to right, you are reading out the word itself. The symbols are written in the order of the sounds, in a straight line (they are linear).

In our writing system, all sounds are arranged in a linear order like this. [actually not all, because there's words like "cradle" and "metre", but we're talking about the basic concept of the script, not the occasional weirdness of English spelling in practice]. A vowel nucleus is written between its onset consonant and the following consonant, whether that's a coda or the onset of the next syllable. Where it's written, its position in the line of symbols on the page, matches when it is said, its position in the stream of sequential sounds people make.

But this isn't necessary. What if we spelled the word A - Y - A - T - A - GH - N?

If we did that, then the CONSONANTS would still be written in their logical sequential order matching the order of speech, but the vowels would not be (relative to the consonants). Vowel nuclei would be written before the consonant onset that actually preceded them in speech.

Or worse, what if we spelled it Y - A - A - T - AN - GH? Again, the onsets would be written in their logical order, but the nuclei and codas would not be. Or not always. In this case, it may be the case that nuclei and codas following Y are written to the right of the onset (the logical order in a left-to-right script), but nuclei and codas following T or GH are actually written to their left.

Or we might imagine the word "sisal" being spelled I - S - S - A - L. Here, the position of the vowel doesn't depend on the nature of the onset but the nature of the vowel: I always precedes its onset, while A always follows it.

Or we might even imagine, though I cangt easily write, a spelling in which the A of sisal was written ABOVE the second S, and maybe the L was written BELOW it.

Again, in all these examples, the ONSETS are being written in a linear order. But the nuclei (and potentially codas) are not. Instead, their position DEPENDS on the 'parent' onset in some way - they are linked together, they're not randomly scattered through the word - rather than independently following the SAME placement rule as the onsets.

One way to see this is what I've said: there are symbols for letters, but the vowels are sometimes written in 'the wrong order', or even above or below the line of consonants rather than in sequence with them. This makes it sound like a badly-ordered alphabet.

The OTHER way to see this is to say that it's not that the ONSETS are in a linear order, but that SYLLABLES are. We can then see above spellings as:

YA-TA-GHA-N
YA-TA-GHAN
SI-SA-L

This makes it sound like a syllabary. Except that in this case the "symbol" for, say, YA, can be "decomposed" into two parts, one of which indicates Y and one of which indicates A. But the rules for how the two parts are positioned on the page relative to one another are governed by a SEPARATE ruleset for how syllable-blocks are composed, NOT by the general rules about how syllables are ORDERED on the page. So it may be that the symbol for YA has the vowel "part" of the symbol precede the consonant part, whereas the symbol for "SI" has the vowel part follow the consonant part, or be written above or below it, for example. Whereas in an alphabet, there is only one ruleset for how symbols are ordered, because vowels and consonants are treated as equivalent and independent, NOT as two parts of the same composite symbol.

Because this system can be seen as an alphabet (with an unusual degree of disordered spelling) or as a syllabary (with an unusual degree of composability in the syllable symbols), it is called an alphasyllabary.

Is that clear now?

[WeepingElf]

Is that clear now?

No

[Salmoneus]

...perhaps, if that's serious and not trolling, you could be more specific!?

Look, you see those letters, right? They form a line. I mean, no, not literally a single line, but the symbols, the letters, are arranged in a row, aren't they? There are two ends of that row, yes? Left, and right.

Now imagine speaking. Your words don't form a line because they spread out in the air. But if you imagine drawing a timeline, a graph of words over time, you would see that your words come one after another. Not simultaneously. Thus forming, as it were, a row of words. You could write your words out in a line, as indeed we do on this form, starting at one side of the line and ending at the other. In English, we write our words from left to right. OK?

So at the left of this line of words you'll find the first words in this sentence, and at the right of it you'll find the last words in this sentence, and the words in the middle are in the order that I say them. OK?

These lines of words represent spoken syllables. Yes?

The order of the spoken syllables, just like the order of the spoken words, can (with one or two oddities of English excepted) be read from left to right. In the case of that last sentence, it started with the syllables THE, OR, DER, OF, THE and SPO, and so on.

The order of the spoken syllables thus matches the order of the written letters. OK?

Specifically, the order of the ONSETS of the syllables, the first consonant in each syllable, is in the order of the written letters. Am I making sense? So in this sentence, SP comes before C, comes before F, comes before C, comes before LL, right? [in English some syllables have null-onsets, let's ignore that for now]

So far, everything is applicable to most writing systems. Now...

In English, the symbolic representation of the SYLLABLE consists of (at least) TWO PARTS. One part represents the ONSET CONSONANT. The other part represents the VOWEL NUCLEUS. The SYLLABLE-SYMBOL can be "decomposed" into two parts - that is, it can be analysed of consisting of two parts. Those two parts can be mixed-and-matched to form new syllables. So, F and I make FI, a syllable. T and U make TU. But we could also have FU, or TI. The pronunciation of these syllable-symbols is (for the most part) deducible because different symbols are systematically related: all the syllable ending U have a certain vowel sound, all the syllables beginning F have a certain consonant sound, and so on (in an ideal spelling system). This is clear, presumably?

Now, let's imagine a writing system in which there was NOT this systematic relationship between syllable-symbols. That is, if I gave you the symbol for FI and the symbol for TU, and explained their meanings, and then showed you the symbol for FU, you would not guess how it was pronounced, because it wouldngt have any systematic, predictable, consistent, understandable graphical relationship to other syllable-symbols that would tell you how it was pronounced. You would have to learn the pronunciation of each syllable one-by-one by rote.

This is what we call a SYLLABIC writing system. It consists of symbols (which can consist of more than one mark on the page!), each of which indicates a specific syllable. You have to learn which symbol represents each syllable.

But note: the syllabic system is STILL LINEAR. The syllables are still ordered in a row from one end (the beginning) to another (the end). [no, it need not always be a straight row, it can be curved, etc, it can even bend back on itself, but it's still a row. There is no ambiguity over which symbol follows which, there is a single one-dimensional sequence. To write "pimento", you start with the syllable PI, then MEN, then TO, in order].

Now let's go back to a system more like ours. In this system, SYLLABLES ARE STILL WRITTEN. In the word "pimento" you can still see the syllables, PI, MEN, and TO, and they are STILL IN ORDER. In a line, as it were.

The difference is, in THIS system, these syllable-symbols are DECOMPOSABLE. The syllable-symbol PI consists of a P, and then an I. OK? P + I = PI. If you know what P indicates and you know what I indicates, you can put them together to work out what PI indicates. You can't do that in a syllabary. But you can in English! (in theory, obviously English spelling can be weird, let's ignore that).

So then the question arises: if we want to write the syllable PI, and we want to do it with the symbol-parts P and I, how do we arrange those parts?

Now, here again we must remember: SYLLABLES IN A WORD ARE WRITTEN IN A LINE. THEY ARE LINEAR.

Therefore each syllable-symbol has a FRONT (closer to the start of the line) and a BACK (closer to the end of the line).

In the case of the syllable PI, we have two phonological parts: an onset consonant P and a nucleus vowel I.

In SPEECH, the nucleus FOLLOWS the onset. And in our line of speech we've said that LATER, i.e. FOLLOWING syllables and words (and sentences, etc) are closer to the BACK of the line. Which in English means closer to the RIGHT.

So, one rule a writing system COULD have is that, WITHIN EACH SYLLABLE, the part symbolising the NUCLEUS is always written at the BACK of the part symbolising the ONSET. Which in a language like English means to the RIGHT of the onset.

So, to make the syllable PI, we don't just randomly place P and I together. We specifically write the I to the RIGHT of the P. OK?

If English were written from right to left - that is, sentences, words and syllables were all read from right to left, we would instead put the I to the LEFT of the P.

That is, in English there is a rule that (in general) the PARTS of a syllable are read from left to right, just as the syllables making up a word are read from left to right and the words making up a sentence are read from left to right, and so on. The construction of syllable-symbols is therefore LINEAR, because it can be read from one side of a line to the other.

We call a system like this an ALPHABET.

In fact, in this system, because each later onset is to the right of the first, AND each nucleus is to the right of its onset, we can read the symbols from left to right, in order, without even having to PAY ATTENTION to the fact that onsets and nuclei make up syllables at all! You could even teach someone (who couldngt read this script) to list the order of sounds in an utterance just by looking at the order of symbols in a line, WITHOUT EVEN TEACHING THEM WHAT CONSONANTS AND VOWELS WERE! In this sense, consonants and vowels in an alphabet have equal status and the distinction doesn't matter (in the abstract, ignoring minor spelling rules in different languages, like how in english doubling a consonant sound shortens the preceding vowel but doubling a vowel doesn't shorten the preceding consonant - but that's just English spelling, it's not true of alphabets in general). You can teach people to speak English without even teaching them what a syllable is!



But syllables don't HAVE to be written this way. We've already talked about writing in a way where syllable-symbols can't be decomposed into vowel and consonant at all! And in that case, you couldn't learn to read the writing without knowing about syllables (at least subconsciously).

And likewise, just ebcause we CAN decompose the syllable-symbol into an onset syllable and a nucleus symbol, THEY DON'T HAVE TO BE ORDERED THE SAME WAY AS IN ENGLISH.

So, to make P and I into a syllable, we could instead write IP. We could write 'pimento' as IP-NEM-OT.

Note that in this example the SYLLABLES SEQUENCE IS STILL READ IN THE SAME DIRECTION. We are not just reversing the direction of the written line as a whole. Later sentences are still to the right of earlier ones. Later words are to the right of earlier ones. Later syllables are to the right of earlier ones. But WITHIN THE SYLLABLE, the LINEAR REPRESENTATION breaks down! Within the syllable, the nucleus, which is later than the onset, is NOT to the right of the onset!

Instead, the two parts, onset and nucleus (and coda in the case of NEM, though a differengt system might write that EM-N, as two syllables, one without a nucleus) must be treated as a BOUND BLOCK. You have to work out how to pronounce that syllable BLOCK, and then sound out each syllable in turn. You cannot start at one end of the line and simply pronounce one sound at a time with no regard for syllables. You have to pronounce one syllable at a time.

This system is like a syllabic system, because you have to work out the pronunciation of each syllable block, and then those blocks are ordered in a line representing the order of spoken speech. Unlike an alphabetic system, the individual sounds -onsets and nuclei - are NOT ordered in a line representing the order of spoken speech.

But this system is like an alphabetic system because there are specific symbols for each sound - both for onsets and for nuclei. It's just that those symbols occur within, are parts of, the syllable-block. Unlike a syllabic system, you CAN work out the pronunciation of each syllable if you know the pronunciation of each part of the symbol.

So this system is in part like an alphabet and in part like a syllabary. So it is called an ALPHASYLLABARY.

OK?

Now, the nuclei in an alphasyllabary don't HAVE to be to the FRONT of the onsets. They just have to NOT be consistently to the BACK. Maybe the symbol for one nucleus is always at the front of the onset symbol, but the symbol for a different nucleus is always at the back. Or maybe one onset is always at the front of the nucleus while another is at the back. Or maybe two certain onsets from adjacent syllables are always adjacent (making a nucleus-onset-onset-nucleus order), while two other onsets don't have to be, or always aren't. The possibilities are endless. And it doesn't matter, because there's never ambiguity which order to read the syllables in: syllables further to the left (in a left-to-right system) are read before symbols to the right. And within the syllable it's clear what order to read the parts: by definition onsets are read first, and then nuclei. So there is no ambiguity.

And sometimes the nuclei and onsets may not even be arranged on a single left-right axis at all! For instance, maybe the onsets/syllables are read left to right, but some nuclei are written ABOVE the onset, and some are written BELOW!

........is that clear now?


If not, could you be specific about what the problem is?

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

And for sake of completeness: the abugida principle, which is unrelated to all of this in theory, is that in some writing systems, a SPECIFIC VALUE OF NUCLEUS is not written at all, and instead an onset written without a nucleus is read as having that nucleus.

Syllabaries cannot be abugidas, because NO nucleus is explicitly and identifiably written in them. However, both alphabets and alphasyllabaries can be, in theory.

In practice, most abugidas are alphasyllabaries, and I think most alphasyllabaries are abugidas. This is partly for historical reasons (both abugidas and alphasyllabaries are often created from abjads through adding vowel diacritics), and partly areal (Indian scripts in general happen to be both).

[WeepingElf]

OK.

[Keenir]

Okay, replying to this before reading your more recent post.

Systematic means having a coherent system.

So, for instance, if you have the syllable-symbol BA and add a mark to make it BAN, and you have a syllable-symbol PA and add the same mark to make it PAN, that's a systematic relationship between symbols.

Okay, that makes sense.

I wasn't trying to be poetic.

It felt poetic; thats not a bad thing.

But this isn't necessary. What if we spelled the word A - Y - A - T - A - GH - N?

If we did that, then the CONSONANTS would still be written in their logical sequential order matching the order of speech, but the vowels would not be (relative to the consonants). Vowel nuclei would be written before the consonant onset that actually preceded them in speech.

But if all the vowels are written that way, then its still coherent, right?

Or we might imagine the word "sisal" being spelled I - S - S - A - L. Here, the position of the vowel doesn't depend on the nature of the onset but the nature of the vowel: I always precedes its onset, while A always follows it.

Okay...but, again, if those are consistent rules regarding A and I, its still coherent. Whereas if SISAL is one of a rare handful of words where the I is written before the S, then its not coherent, yes?

The OTHER way to see this is to say that it's not that the ONSETS are in a linear order, but that SYLLABLES are. We can then see above spellings as:

YA-TA-GHA-N
YA-TA-GHAN
SI-SA-L

This makes it sound like a syllabary. Except that in this case the "symbol" for, say, YA, can be "decomposed" into two parts, one of which indicates Y and one of which indicates A. But the rules for how the two parts are positioned on the page relative to one another are governed by a SEPARATE ruleset for how syllable-blocks are composed, NOT by the general rules about how syllables are ORDERED on the page. So it may be that the symbol for YA has the vowel "part" of the symbol precede the consonant part, whereas the symbol for "SI" has the vowel part follow the consonant part, or be written above or below it, for example.

so the sign which is used for words which have "si" in them, use a sign which is in fact IS.

And that aspect, the use of IS for words with "si", is sufficient to stop it from being a syllabary? Interesting.

Is that clear now?

Its more clear than things were before now, yes.

I'll read your next post later tonight.

[Keenir]

In English, the symbolic representation of the SYLLABLE consists of (at least) TWO PARTS. One part represents the ONSET CONSONANT. The other part represents the VOWEL NUCLEUS. The SYLLABLE-SYMBOL can be "decomposed" into two parts - that is, it can be analysed of consisting of two parts. Those two parts can be mixed-and-matched to form new syllables. So, F and I make FI, a syllable. T and U make TU. But we could also have FU, or TI. The pronunciation of these syllable-symbols is (for the most part) deducible because different symbols are systematically related: all the syllable ending U have a certain vowel sound, all the syllables beginning F have a certain consonant sound, and so on (in an ideal spelling system). This is clear, presumably?

so, like Hindi?

Now, let's imagine a writing system in which there was NOT this systematic relationship between syllable-symbols. That is, if I gave you the symbol for FI and the symbol for TU, and explained their meanings, and then showed you the symbol for FU, you would not guess how it was pronounced, because it wouldngt have any systematic, predictable, consistent, understandable graphical relationship to other syllable-symbols that would tell you how it was pronounced. You would have to learn the pronunciation of each syllable one-by-one by rote.

So, more like Cherokee?

This is what we call a SYLLABIC writing system. It consists of symbols (which can consist of more than one mark on the page!), each of which indicates a specific syllable. You have to learn which symbol represents each syllable.

Now let's go back to a system more like ours. In this system, SYLLABLES ARE STILL WRITTEN. In the word "pimento" you can still see the syllables, PI, MEN, and TO, and they are STILL IN ORDER. In a line, as it were.

The difference is, in THIS system, these syllable-symbols are DECOMPOSABLE. The syllable-symbol PI consists of a P, and then an I. OK? P + I = PI. If you know what P indicates and you know what I indicates, you can put them together to work out what PI indicates. You can't do that in a syllabary.

But that presupposes that we start out knowing what P and what I indicate, thus we know what PI indicates when using an alphabet...but we don't start out knowing what PI indicates?

So, one rule a writing system COULD have is that, WITHIN EACH SYLLABLE, the part symbolising the NUCLEUS is always written at the BACK of the part symbolising the ONSET. Which in a language like English means to the RIGHT of the onset.

So, to make the syllable PI, we don't just randomly place P and I together. We specifically write the I to the RIGHT of the P. OK?

If we're making our own system, rather than adapting a system we picked up from somewhere/one else.

And likewise, just ebcause we CAN decompose the syllable-symbol into an onset syllable and a nucleus symbol, THEY DON'T HAVE TO BE ORDERED THE SAME WAY AS IN ENGLISH.

So, to make P and I into a syllable, we could instead write IP. We could write 'pimento' as IP-NEM-OT.

Instead, the two parts, onset and nucleus (and coda in the case of NEM, though a differengt system might write that EM-N, as two syllables, one without a nucleus) must be treated as a BOUND BLOCK. You have to work out how to pronounce that syllable BLOCK, and then sound out each syllable in turn. You cannot start at one end of the line and simply pronounce one sound at a time with no regard for syllables. You have to pronounce one syllable at a time.

When we're learning it, yes; after that, it'd be as second nature as pronouncing "knick".

Now, the nuclei in an alphasyllabary don't HAVE to be to the FRONT of the onsets. They just have to NOT be consistently to the BACK. Maybe the symbol for one nucleus is always at the front of the onset symbol, but the symbol for a different nucleus is always at the back.

wait...whats the difference between "consistently" and "always" in this case - it reads like you're contradicting yourself.

And for sake of completeness: the abugida principle, which is unrelated to all of this in theory, is that in some writing systems, a SPECIFIC VALUE OF NUCLEUS is not written at all, and instead an onset written without a nucleus is read as having that nucleus.

Syllabaries cannot be abugidas, because NO nucleus is explicitly and identifiably written in them. However, both alphabets and alphasyllabaries can be, in theory.

wait...but if there is no nucleus in a CV, then why do alphabets count as having a nucleus? surely they too are only (at least in theory) purely onset because they are (at least in principle) one sign = one sound, aka either C or V. {nevermind that; my bad}

wait...they aren't abugidas because something isn't written in it? But the people who read those syllabaries still speak the nucleus, right?

[regenbogen9]

...perhaps, if that's serious and not trolling, you could be more specific!?

Look, you see those letters, right? They form a line. I mean, no, not literally a single line, but the symbols, the letters, are arranged in a row, aren't they? There are two ends of that row, yes? Left, and right.

Now imagine speaking. Your words don't form a line because they spread out in the air. But if you imagine drawing a timeline, a graph of words over time, you would see that your words come one after another. Not simultaneously. Thus forming, as it were, a row of words. You could write your words out in a line, as indeed we do on this form, starting at one side of the line and ending at the other. In English, we write our words from left to right. OK?

So at the left of this line of words you'll find the first words in this sentence, and at the right of it you'll find the last words in this sentence, and the words in the middle are in the order that I say them. OK?

These lines of words represent spoken syllables. Yes?

The order of the spoken syllables, just like the order of the spoken words, can (with one or two oddities of English excepted) be read from left to right. In the case of that last sentence, it started with the syllables THE, OR, DER, OF, THE and SPO, and so on.

The order of the spoken syllables thus matches the order of the written letters. OK?

Specifically, the order of the ONSETS of the syllables, the first consonant in each syllable, is in the order of the written letters. Am I making sense? So in this sentence, SP comes before C, comes before F, comes before C, comes before LL, right? [in English some syllables have null-onsets, let's ignore that for now]

So far, everything is applicable to most writing systems. Now...

In English, the symbolic representation of the SYLLABLE consists of (at least) TWO PARTS. One part represents the ONSET CONSONANT. The other part represents the VOWEL NUCLEUS. The SYLLABLE-SYMBOL can be "decomposed" into two parts - that is, it can be analysed of consisting of two parts. Those two parts can be mixed-and-matched to form new syllables. So, F and I make FI, a syllable. T and U make TU. But we could also have FU, or TI. The pronunciation of these syllable-symbols is (for the most part) deducible because different symbols are systematically related: all the syllable ending U have a certain vowel sound, all the syllables beginning F have a certain consonant sound, and so on (in an ideal spelling system). This is clear, presumably?

Now, let's imagine a writing system in which there was NOT this systematic relationship between syllable-symbols. That is, if I gave you the symbol for FI and the symbol for TU, and explained their meanings, and then showed you the symbol for FU, you would not guess how it was pronounced, because it wouldngt have any systematic, predictable, consistent, understandable graphical relationship to other syllable-symbols that would tell you how it was pronounced. You would have to learn the pronunciation of each syllable one-by-one by rote.

This is what we call a SYLLABIC writing system. It consists of symbols (which can consist of more than one mark on the page!), each of which indicates a specific syllable. You have to learn which symbol represents each syllable.

But note: the syllabic system is STILL LINEAR. The syllables are still ordered in a row from one end (the beginning) to another (the end). [no, it need not always be a straight row, it can be curved, etc, it can even bend back on itself, but it's still a row. There is no ambiguity over which symbol follows which, there is a single one-dimensional sequence. To write "pimento", you start with the syllable PI, then MEN, then TO, in order].

Now let's go back to a system more like ours. In this system, SYLLABLES ARE STILL WRITTEN. In the word "pimento" you can still see the syllables, PI, MEN, and TO, and they are STILL IN ORDER. In a line, as it were.

The difference is, in THIS system, these syllable-symbols are DECOMPOSABLE. The syllable-symbol PI consists of a P, and then an I. OK? P + I = PI. If you know what P indicates and you know what I indicates, you can put them together to work out what PI indicates. You can't do that in a syllabary. But you can in English! (in theory, obviously English spelling can be weird, let's ignore that).

So then the question arises: if we want to write the syllable PI, and we want to do it with the symbol-parts P and I, how do we arrange those parts?

Now, here again we must remember: SYLLABLES IN A WORD ARE WRITTEN IN A LINE. THEY ARE LINEAR.

Therefore each syllable-symbol has a FRONT (closer to the start of the line) and a BACK (closer to the end of the line).

In the case of the syllable PI, we have two phonological parts: an onset consonant P and a nucleus vowel I.

In SPEECH, the nucleus FOLLOWS the onset. And in our line of speech we've said that LATER, i.e. FOLLOWING syllables and words (and sentences, etc) are closer to the BACK of the line. Which in English means closer to the RIGHT.

So, one rule a writing system COULD have is that, WITHIN EACH SYLLABLE, the part symbolising the NUCLEUS is always written at the BACK of the part symbolising the ONSET. Which in a language like English means to the RIGHT of the onset.

So, to make the syllable PI, we don't just randomly place P and I together. We specifically write the I to the RIGHT of the P. OK?

If English were written from right to left - that is, sentences, words and syllables were all read from right to left, we would instead put the I to the LEFT of the P.

That is, in English there is a rule that (in general) the PARTS of a syllable are read from left to right, just as the syllables making up a word are read from left to right and the words making up a sentence are read from left to right, and so on. The construction of syllable-symbols is therefore LINEAR, because it can be read from one side of a line to the other.

We call a system like this an ALPHABET.

In fact, in this system, because each later onset is to the right of the first, AND each nucleus is to the right of its onset, we can read the symbols from left to right, in order, without even having to PAY ATTENTION to the fact that onsets and nuclei make up syllables at all! You could even teach someone (who couldngt read this script) to list the order of sounds in an utterance just by looking at the order of symbols in a line, WITHOUT EVEN TEACHING THEM WHAT CONSONANTS AND VOWELS WERE! In this sense, consonants and vowels in an alphabet have equal status and the distinction doesn't matter (in the abstract, ignoring minor spelling rules in different languages, like how in english doubling a consonant sound shortens the preceding vowel but doubling a vowel doesn't shorten the preceding consonant - but that's just English spelling, it's not true of alphabets in general). You can teach people to speak English without even teaching them what a syllable is!



But syllables don't HAVE to be written this way. We've already talked about writing in a way where syllable-symbols can't be decomposed into vowel and consonant at all! And in that case, you couldn't learn to read the writing without knowing about syllables (at least subconsciously).

And likewise, just ebcause we CAN decompose the syllable-symbol into an onset syllable and a nucleus symbol, THEY DON'T HAVE TO BE ORDERED THE SAME WAY AS IN ENGLISH.

So, to make P and I into a syllable, we could instead write IP. We could write 'pimento' as IP-NEM-OT.

Note that in this example the SYLLABLES SEQUENCE IS STILL READ IN THE SAME DIRECTION. We are not just reversing the direction of the written line as a whole. Later sentences are still to the right of earlier ones. Later words are to the right of earlier ones. Later syllables are to the right of earlier ones. But WITHIN THE SYLLABLE, the LINEAR REPRESENTATION breaks down! Within the syllable, the nucleus, which is later than the onset, is NOT to the right of the onset!

Instead, the two parts, onset and nucleus (and coda in the case of NEM, though a differengt system might write that EM-N, as two syllables, one without a nucleus) must be treated as a BOUND BLOCK. You have to work out how to pronounce that syllable BLOCK, and then sound out each syllable in turn. You cannot start at one end of the line and simply pronounce one sound at a time with no regard for syllables. You have to pronounce one syllable at a time.

This system is like a syllabic system, because you have to work out the pronunciation of each syllable block, and then those blocks are ordered in a line representing the order of spoken speech. Unlike an alphabetic system, the individual sounds -onsets and nuclei - are NOT ordered in a line representing the order of spoken speech.

But this system is like an alphabetic system because there are specific symbols for each sound - both for onsets and for nuclei. It's just that those symbols occur within, are parts of, the syllable-block. Unlike a syllabic system, you CAN work out the pronunciation of each syllable if you know the pronunciation of each part of the symbol.

So this system is in part like an alphabet and in part like a syllabary. So it is called an ALPHASYLLABARY.

OK?

Now, the nuclei in an alphasyllabary don't HAVE to be to the FRONT of the onsets. They just have to NOT be consistently to the BACK. Maybe the symbol for one nucleus is always at the front of the onset symbol, but the symbol for a different nucleus is always at the back. Or maybe one onset is always at the front of the nucleus while another is at the back. Or maybe two certain onsets from adjacent syllables are always adjacent (making a nucleus-onset-onset-nucleus order), while two other onsets don't have to be, or always aren't. The possibilities are endless. And it doesn't matter, because there's never ambiguity which order to read the syllables in: syllables further to the left (in a left-to-right system) are read before symbols to the right. And within the syllable it's clear what order to read the parts: by definition onsets are read first, and then nuclei. So there is no ambiguity.

And sometimes the nuclei and onsets may not even be arranged on a single left-right axis at all! For instance, maybe the onsets/syllables are read left to right, but some nuclei are written ABOVE the onset, and some are written BELOW!

........is that clear now?


If not, could you be specific about what the problem is?

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

And for sake of completeness: the abugida principle, which is unrelated to all of this in theory, is that in some writing systems, a SPECIFIC VALUE OF NUCLEUS is not written at all, and instead an onset written without a nucleus is read as having that nucleus.

Syllabaries cannot be abugidas, because NO nucleus is explicitly and identifiably written in them. However, both alphabets and alphasyllabaries can be, in theory.

In practice, most abugidas are alphasyllabaries, and I think most alphasyllabaries are abugidas. This is partly for historical reasons (both abugidas and alphasyllabaries are often created from abjads through adding vowel diacritics), and partly areal (Indian scripts in general happen to be both).

Thanks very much for the explanations!
So if in a language there is separate symbols for both consonants and vowels, but vowels are written as diacritics on top or below consonants (maybe similars to the independent vowel symbol) it is what? And if there is an inherent vowel for consonant + a and the rest vowels change according to diacritics, it changes the type of writing system?
And I think that if you give an example of a natural abugida, a natural alphasyllabary, a natural syllabary... It would help a lot.

[Salmoneus]

I'm sorry if my last post sounded aggressive; I'm just frustrated with myself, because after writing hundreds and hundred of words to explain something that to me seems incredibly obvious and originates in about two sentences on wikipedia, evidently I'm still speaking gibberish that nobody can understand!

In English, the symbolic representation of the SYLLABLE consists of (at least) TWO PARTS. One part represents the ONSET CONSONANT. The other part represents the VOWEL NUCLEUS. The SYLLABLE-SYMBOL can be "decomposed" into two parts - that is, it can be analysed of consisting of two parts. Those two parts can be mixed-and-matched to form new syllables. So, F and I make FI, a syllable. T and U make TU. But we could also have FU, or TI. The pronunciation of these syllable-symbols is (for the most part) deducible because different symbols are systematically related: all the syllable ending U have a certain vowel sound, all the syllables beginning F have a certain consonant sound, and so on (in an ideal spelling system). This is clear, presumably?

so, like Hindi?

Or English! This is how English works! I gave examples from English!

Now, let's imagine a writing system in which there was NOT this systematic relationship between syllable-symbols. That is, if I gave you the symbol for FI and the symbol for TU, and explained their meanings, and then showed you the symbol for FU, you would not guess how it was pronounced, because it wouldngt have any systematic, predictable, consistent, understandable graphical relationship to other syllable-symbols that would tell you how it was pronounced. You would have to learn the pronunciation of each syllable one-by-one by rote.

So, more like Cherokee?

So far as I can see, yes.

This is what we call a SYLLABIC writing system. It consists of symbols (which can consist of more than one mark on the page!), each of which indicates a specific syllable. You have to learn which symbol represents each syllable.

Now let's go back to a system more like ours. In this system, SYLLABLES ARE STILL WRITTEN. In the word "pimento" you can still see the syllables, PI, MEN, and TO, and they are STILL IN ORDER. In a line, as it were.

The difference is, in THIS system, these syllable-symbols are DECOMPOSABLE. The syllable-symbol PI consists of a P, and then an I. OK? P + I = PI. If you know what P indicates and you know what I indicates, you can put them together to work out what PI indicates. You can't do that in a syllabary.

But that presupposes that we start out knowing what P and what I indicate, thus we know what PI indicates when using an alphabet...but we don't start out knowing what PI indicates?

Errr... huh? I'm not claiming that any one writing system is innate, if that's what you mean. All writing systems have to be learnt - in an alphabet you don't start off knowing what P and I mean, and in a syllabary you don't start out knowing what PI means. You have to learn these things. But with an alphabet, once you learn P and I, and other phoneme symbols (probably only a few dozen) you know what PI means, and all the hundreds of other syllable combinations formed from those elements. Whereas in a syllabary you have to learn all of the hundreds of syllable symbols independently.

So, one rule a writing system COULD have is that, WITHIN EACH SYLLABLE, the part symbolising the NUCLEUS is always written at the BACK of the part symbolising the ONSET. Which in a language like English means to the RIGHT of the onset.

So, to make the syllable PI, we don't just randomly place P and I together. We specifically write the I to the RIGHT of the P. OK?

If we're making our own system, rather than adapting a system we picked up from somewhere/one else.

What? No! No, when you learn English, that's how you write! You are taught to write from left to right? Aren't you? Is this just me? Am I going insane!?

And likewise, just ebcause we CAN decompose the syllable-symbol into an onset syllable and a nucleus symbol, THEY DON'T HAVE TO BE ORDERED THE SAME WAY AS IN ENGLISH.

So, to make P and I into a syllable, we could instead write IP. We could write 'pimento' as IP-NEM-OT.

Instead, the two parts, onset and nucleus (and coda in the case of NEM, though a differengt system might write that EM-N, as two syllables, one without a nucleus) must be treated as a BOUND BLOCK. You have to work out how to pronounce that syllable BLOCK, and then sound out each syllable in turn. You cannot start at one end of the line and simply pronounce one sound at a time with no regard for syllables. You have to pronounce one syllable at a time.

When we're learning it, yes; after that, it'd be as second nature as pronouncing "knick".

I didn't say it wasn't EASY or NATURAL. I just said what you had to do. You have to start at one side of the line and sound out the syllables one by one. Just as in English you are taught to sound out the phonemes one by one [allowing for some irregular spellings in English]

Now, the nuclei in an alphasyllabary don't HAVE to be to the FRONT of the onsets. They just have to NOT be consistently to the BACK. Maybe the symbol for one nucleus is always at the front of the onset symbol, but the symbol for a different nucleus is always at the back.

wait...whats the difference between "consistently" and "always" in this case - it reads like you're contradicting yourself.

There is a difference between "X doesn't to always be Y" and "X has to not always be Y". Isn't there?

And for sake of completeness: the abugida principle, which is unrelated to all of this in theory, is that in some writing systems, a SPECIFIC VALUE OF NUCLEUS is not written at all, and instead an onset written without a nucleus is read as having that nucleus.

Syllabaries cannot be abugidas, because NO nucleus is explicitly and identifiably written in them. However, both alphabets and alphasyllabaries can be, in theory.

wait...but if there is no nucleus in a CV, then why do alphabets count as having a nucleus? surely they too are only (at least in theory) purely onset because they are (at least in principle) one sign = one sound, aka either C or V. {nevermind that; my bad}

wait...they aren't abugidas because something isn't written in it? But the people who read those syllabaries still speak the nucleus, right?

I don't... I don't understand your confusion here.

Abugidas don't write certain vowels.

Syllabaries have a distinct symbol for each syllable.

Yes, people who speak LANGUAGES that are sometimes written in SYLLABARIES do still pronounce vowels. Why wouldn't they!?

A speaker of Cherokee is not going to miss out all the vowels when reading Cherokee written in a syllabary, yet pronounce them when they read Cherokee written in the Latin alphabet!

We're talking about WRITING here, NOT SPEECH. Any language can be written with any sort of writing system (though some may be less congenial matches than others).