In recent posts, I developed a way to write Sanskrit using the Thaana script, determined the sounds that constitute “true Dhivehi”, and explained various ways to expand the Dhivehi vocabulary. This post will draw on the ideas from all three of those previous posts. We will see how Sanskrit evolved into Dhivehi, looking specifically at the way words changed. Using this, we will determine the nature of phonological changes between the two languages which we will then use as the basis for the naturalisation of new Sanskrit words into Dhivehi.
First, allow me to explain the analogy in the title of the post.
It’s pretty simple to understand. For one language to evolve into another, it passes through a “filter”. This filter adds, removes, or changes the sounds of the original language in regular, predictable ways, to leave you with the new language, much in the same way that a sieve regularly and predictably removes seeds and pulp from freshly squeezed lemon juice. And for this reason, I shall dub the Dhivehi name for this phenomenon “އަޑުފުރޭނި” (literally “sound-sieve”).
In this post, we are aiming to find out the nature of the phonological filter that turns Sanskrit into Dhivehi. What sounds does it preserve? What sounds does it change? And in what way? Below is a flow chart that represents the phonological filter. We are trying to find out what happens in the second box.
Let us begin!
Uncovering the Filter
To determine the nature of the filter, we have to look at words before and after they pass through it. This simply meaning examining Dhivehi words and their Sanskrit etymons, and looking for recurring patterns.
The more words examined, the more accurately we can know how the filter behaves. But to keep things orderly, I’ll group the words by topic. To represent the Sanskrit words, I will use the system I devised in this post. It’s pretty straightforward, but you might want to familiarise yourself with it a little before continuing.
This first topic is a bit out there, but I think it’s very interesting. The Dhivehi words are probably more representative of an older form of the language, but that simply gives us a more accurate representation of sound changes. First, here are the 27 lunar mansions of the Hindu/Dhivehi astrological calendar:
And here are the zodiac signs, or the solar months:
Just by looking at these words, there are already some rules we can deduce about the sound changes necessary to go from Sanskrit to Dhivehi. These include:
- Loss of consonant clusters. In cases where a “strong” consonant precedes a “weak” one, the “strong” consonant, typically a stop consonant, will remain in the Dhivehi word, while the “weak” one is eliminated. For example, the /ށ/ and /ރ/ are lost from the word ނަކްށަތްރަ, while the stop consonants /ކ/ and /ތ/ remain, leaving ނަކަތް. If the “weak” consonant is first, a vowel may be inserted, for example ކަރްކަޓަ becomes ކަރުކަށަ.
- The consonant cluster /ށްޓ/ becomes /ށ/. Historically, however, there was probably in intermediate phase during which the /ޓ/ remained by itself before it changed to /ށ/.
- Loss of aspirated consonants. Instead, these consonants take their non-aspirated forms in the Dhivehi words
- Sanskrit /ޕ/ becomes Dhivehi /ފ/
- Loss of distinction between /ޱ/ and /ނ/
- Sanskrit /ޑހ/ becomes Dhivehi /ޅ/
- Sanskrit /ޖ/ becomes Dhivehi /ދ/
- The syllabic consonant /ރ/ either becomes a short vowel as in ކެތި/ކރތްތިކާ or becomes the onset of a new syllable with the addition of vowels, as in ވިރުސިކަ/ވރޝްޗިކަ
- Sanskrit /ޝ/ ,/ށ/ and /ޗ/ become Dhivehi /ހ/ or less commonly /ސ/
Now let’s look at more words, which can help to confirm these rules, or give us more details about how they work.
Apart from showing us sound changes, these words also show us changes in meaning and usage. For example, while އަކްށި means “eye”, the Dhivehi descendant އެސް is rarely used by itself to mean “eye”, instead it is most commonly used in compounds like އެސްފިޔަ (“eyelash”). The word ކޭޝަ refers to hair in general, while the Dhivehi word ކެސް refers only to pubic hair. Meanwhile, ކަޱްޓަ, the etymon of ކަށި means “thorn” or “nib”, instead of “bone”.
Many of the sound-change rules from the previous sets of words are confirmed here:
- /ޕ/ changes to /ފ/
- /ޱ/ merges with /ނ/
- /ޖ/ changes to /ދ/
- The syllabic /ރ/ becomes a short vowel
- Consonant clusters are reduced
There are also some new patterns which emerge from these words:
- Nasal codas become a prenasalised onset if the following syllable begins with a voiced consonant, or are removed entirely if the following syllable begins with an unvoiced consonant
- When /ވ/ is the second consonant in a cluster, it can be reduced to the vowel /އޫ/
- If a word ends with /އަ/ (which would be closer to a schwa in Sanskrit), it is not pronounced and the preceding consonant becomes the coda of the preceding syllable, for sounds where this is permissible according to Dhivehi phonotactics. This sound change probably occurred at a later stage in the evolution of Dhivehi. For example, you might have noticed that ހަސްތަ became އަތަ in the context of lunar mansions, but became އަތް in the context of body parts. Many other astrology words end with /އަ/ as well, the likely explanation being that they were “frozen in time” due to the limited contexts in which they are used. It is relatively uncommon for modern Dhivehi words to end in /އަ/ and the schwa deletion rule explains this.
Once again, these words show us how meanings and usage change. I won’t go into the details here, because I think the etymologies of colours are interesting enough to warrant their own post.
I am not sure about ޕަތްރަ being the etymon of ފެހި. If this is true, I believe it would have come into Dhivehi through the Dravidian languages. For example, in Malayalam, the word is ޕައްޗަ, and in Tamil, it is ޕައްޗައި or ޕަސުމް. Otherwise, I don’t have an explanation for the sound change.
Most of the changes here confirm the patterns we have already seen. One of the stranger ones is the change from ނީލަ to ނޫ. I believe this would have occurred due to the /ލ/ being replaces with a rounded back vowel which then “took over” the initial /އި/. So the entire process may have gone like this: ނީލަ, then ނިލް, then ނިއު, and finally ނޫ. Such a change is not without precedent in Dhivehi, and this hypothesis becomes more plausible when you consider the following:
- In some southern dialects, which are more conservative, Laamu sukun (ލް) is phonotactically valid
- Southern dialects often use /އި/ where the Malé dialect would use /އު/, for example, ދީނި instead of ދޫނި
The point I’m trying to make is that Sanskrit words ending with /ލ/ can cause unexpected vowel changes in Dhivehi descendants.
These words confirm the patterns we have already found with regards to consonants, but they also demonstrate that vowels behave somewhat unpredictably. Perhaps there is method to the madness.
Consider the word ދަކްށިޱަ. If we apply the necessary consonant change patterns we have found so far (that is, changing ޱ to ނ and removing the weak consonant ށ from the cluster) the word we end up with is ދަކިނަ. So why, then, do the vowels change? I believe it may have something to do with vowel harmony. /އަ/ is an open front vowel, and /އި/ is a closed front vowel. Alternating them in rapid succession in a word like ދަކިނަ is somewhat taxing. Because emphasis is on the first syllable ދަ, the /އި/ in ކި is not pronounced as clearly, and this vowel is prone to change. ކި changes to ކު because /ކ/ is a velar consonant pronounced at the back of the mouth and /އު/ is a closed back vowel, so these sounds kind of go together naturally. ނަ changes to ނު due to vowel harmony (that is, to be the same as ކު), and ދަ changes to ދެ in order to “balance out” the vowels. /އެ/ is a front vowel like /އަ/ but it is a close-mid vowel, meaning it’s closer to /އު/.
That’s my long-winded post-hoc explanation. I have no idea how valid it is. But I think the best “rule” for determining how vowels should change is to simplify as much as possible, and make words as easy to say as possible. This applies to consonants too. And when you look at how some of the other words in the list above have changed, you can see how this rule applies.
Revealing the Filter
Now that we have found the patterns, we can combine them all to show how the Sanskrit-Dhivehi phonological filter works. Here is the same flow chart from before, with the middle section filled in:
Note that this filter is still incomplete. I may have missed some rules, and I didn’t go into full details about how each one is applied. Some of these rules don’t even seem to apply consistently. In reality, sound changes that have occurred between Sanskrit and Dhivehi are much more complicated. But these rules give us a good idea of how Sanskrit words can be naturalised, which brings us to the next section:
Using the Filter
Because we now have some idea of how to make Sanskrit words Dhivehi, we can take Sanskrit words that have not made it into Dhivehi, and force them through the filter to expand the Dhivehi vocabulary. I’ll give you a few examples, going step by step to show how the sound change rules apply.
1. ކްށޭތްރަ – “Field”
This word has two consonant clusters. So let’s only keep the strong consonants:
Now we get rid of the final vowel (which would be a schwa in Sanskrit), which gives us our final word:
ކޭތަ is actually a Dhivehi word, meaning “eclipse”, but it has a different etymology.
2. ބހްރާތރ – “Brother”
You can probably tell that this word is related to the English word. Let’s start the changes by getting rid of the aspirated consonant:
Now, we can get rid of the consonant cluster, keeping only the strong consonant:
We are left with the syllabic consonant. We can either change it into a short vowel, or insert vowels, while maintaining vowel harmony. This gives us two possible options for the final word:
ބާތި or ބާތަރު
With the second of these two options, you can even go a step further and add voicing to the middle consonant:
This keeps things consistent with existing Dhivehi words like މާދަރީ. So in this case, ބާދަރީ would mean “brotherly”. However, ބާދަރު might sound funny to some people because it also means “old wood”.
3. ގރށްޓި – “Young female animal”
The meaning of this word mostly applies to cows. Amongst its other meanings, one of the more interesting ones is “a woman with only one child”. To change it to Dhivehi, we can start by changing the syllabic consonant into a vowel:
Next, we keep only the string consonant in the cluster:
Because that strong consonant was an unvoiced retroflex stop, we have to change it to a retroflex sibilant, giving us our final word:
This word would behave like a normal Dhivehi word ending with ށި when adding suffixes. So it would be ގިއްޓަށް ,ގިއްޓެއް etc.
This word should also give you an idea of how Sanskrit can enrich Dhivehi. We now have a word to describe the phenomenon that is women with only one child. Sure, it’s not going to be used terribly often, bot neither are the coconut words that people sometimes harp on about.
I’ll also point out now that these new words aren’t just made up out of nowhere, which is often the perception when new words are introduced to the Maldivian public. They are derived by repeating a process that has occurred over centuries. These words are just as valid and just as Dhivehi as the words people use every day.
4. ޝޭށަ – “Remains”
This is another word with several meanings. Others include “waste”, “keepsake” and “survivor”; all of which are related to the idea of being left behind. This word has two sibilants. We can start the changes by changing them into glottal fricatives:
Now we just have to get rid of the final schwa. There are two ways this can be done:
ހޭހި or ހޭހު
The proper from of ހޭހު in Dhivehi would actually be ހޭސް, since /ސ/ and /ހ/ have a somewhat allophonic relationship. ހޭހި will also regain a sibilant with suffixes are added: ހޭއްސަށް ,ހޭއްސެއް etc.
5. ޕޯޑު – “Top of skull” (frontal and parietal bones)
This word requires two sound changes to become Dhivehi. First, the bilabial stop must be changed into a labio-dental fricative:
Next, the retroflex stop must be changed into a retroflex lateral approximant:
Once again, this word shows the potential of new Sanskrit words to expand the Dhivehi vocabulary. This word would be useful specifically for the medical vocabulary of Dhivehi. You could even specify bones by adding prefixes. For example, the frontal bone could be ކުރިފޯޅު.
Knowing the phonological filter that separates Sanskrit and Dhivehi facilitates the naturalisation of Sanskrit words which in turn helps to expand the Dhivehi vocabulary. Having a set of rules to follow when changing the sounds of words makes the process of making new words faster, while ensuring that the words still sound Dhivehi. The next step is to actually put this into practice and start importing vocabulary. I might dedicate an entire post to this in the future.