Accusative form for nouns ending in -a


In the Pali primer by Lily de Silva, the accusative form for nouns ending in -a is ‘+ ŋ’, but I don’t see this form in the suttas in SC, which form is it used in this cases?

Some examples in the primer are:

  1. nara + ŋ = naraŋ
  2. mātula + ŋ = mātulaŋ
  3. kassaka + ŋ = kassakaŋ

Thank you.

1 Like

Suttacentral uses ṃ character for this sound, also although theoretically ṅ is different letter (but the same sound) it’s sometimes used interchangably with ṃ in the middle of the words.


Thank you very much.

I have been trying to translate this word: sambahulānaṃ
But the online translator does not seem to find it. The closes I found was the word ‘sambahula’, but this leaves ‘naṃ’ as a suffix, not just ṃ.

Is sambahula the correct root for the word sambahulānaṃ?

Nerd mode ON

Well, actually, they are different sounds. ṃ is pronounced as a mixture of nasalized m and w (just move your lips as if you would like to pronounce the m sounds without actually doing it, and nasalize the sound all the way), whereas the ṅ sound is good old ng as in the English -ing ending. Quite often the ṃ diachritic (anusvara) was originally used as an anunasika, i.e. as a sign of nasalization of a preceding vowel (see Oberlies’ Pali Grammar) - and there is a world of difference between a nasalized vowel and a vowel plus nasalized sonorant. Where exactly the ṃ diachritic was used as an anusvara and where as an anunasika is a complicated matter, just as it would be difficult to find out how to pornounce the English gh character combination as in ‘enough’ or ‘drought’ if all English native spakers disappeared.
The ṃ and ṅ sounds are pronounced identically in the Singhalese chanting tradition and possibly elsewhere in the modern Theravada world, but in actuality they used to be different sounds.

Nerd mode OFF


Yes, sambahula is the correct root. Sambahulānaṃ is genitive / dative plural.

1 Like

Of course that’s what I actually meant, that it’s spelled the same nowadays :wink:
In Polish we have three such sounds, that once were pronounced differently… now they’re the main source of fear and dread in primary school dictation exercises :stuck_out_tongue:

1 Like

Thanks, that makes sense.

Lets see if I got this correctly then.

In the Pali primer this is how they are inflected:

  1. nara + ŋ = naraŋ
  2. mātula + ŋ = mātulaŋ
  3. kassaka + ŋ = kassakaŋ


  1. nara + e = nare
  2. mātula + e = mātule
  3. kassaka + e = kassake

But in this form, sambahula would be:
Singular: sambahula
Plural: sambahulānaṃ

Or did I get lost?

1 Like

You got lost, but don’t worry :wink:
sambahulānaṃ would be dative / genitive plural, you are inflecting accusative here, so in plural it will be: sambahule.

The tables I have also give possible ending of -āni, but honestly I don’t know how / when it is used, and if it ever shows up as sambahulāni, or maybe only with some irregular words.


Also, as a side note I’m not sure if sambahula can be treated as singular at all, since it means “many”, so is plural by definition.

1 Like

You pretty much did. The ŋ character reflects the modern pronunciation of the ṃ diachritic that Eisel Mazard, the guy who published this Pali primer online, if I am not mistaken, for some reason decided to use instead of the regular anusvara character. Substitue ŋ with ṃ and you will have the PTS spelling of the corresponding forms, i.e.


  1. naraṃ
  2. mātulaṃ
  3. kassakaṃ


  1. nare
  2. mātule
  3. kassake

The -ānaṃ forms are actually Gen. Pl. and Dat. Pl.:

  1. narānaṃ
  2. mātulānaṃ
  3. kassakānaṃ

Cf. the Iti pi so gatha:

satthā devamanussānaṃ ‘teacher of gods and men

Generally, I would recommend to ignore the Pali primers and start with the Warder textbook. It is rather dense and difficult for someone without much linguistic knowledge, but it is still orders of magnitude better.


It look too good to be true :slight_smile:

I need to review the whole dative, genitive, accusative an so on stuff, and I thought after high school I was safe from that!

That makes sense.

Yes! I found that in the Primer now.

I will look for the other textbook that you recommended, I can use all the help and to be honest.

According to Bhante Sujato:

ṁ is correct, ṃ is a vile abomination

But he was outvoted when SuttaCentral started … :laughing:


It is nice to see that democracy works in buddhism.


And just to clarify (or confuse), the ŋ form was used occasionally in the very early days of transcription, but was replaced by the end of the 19th century. It is not part of any contemporary standard or set of conventions, and I don’t really know why anyone would use it today.

Or should we see it as evidence that democracy, in fact, does not work?


I can’t remember which sociologist said: in a world where the majority rules, who protects the minorities?

I agree, those forms look very weird and they actually complicate learning the language. I would vote for an even simpler alphabet.

I am getting there, getting used to all the new things and variations. I have thousands of questions, so you will see me here often.

Thanks for the link, I will look into it.


Abomination may be a little strong, but I never understood the logic of using ṃ.

ṁ/ṃ is similar to ṅ (and commonly pronounced identically) but is quite different from ṇ.

But I’m sure I’m missing some really good reason for the choice, and I see that the Wisdom publications use ṃ…

1 Like

Okay, so I had my suppositions about this, but I decided to solve the problem once and for all with some font forensics. Check this out:

These are scanned images of the original PTS editions, which are still the standard Pali editions in academia. The one on the right is the Majjhima.

As you can see, SN, AN, and DN use the ṃ, while MN uses the ṁ. So we can see that this schism is very old!

Okay, so what’s the difference? Look closer, it’s all in the J. The fonts used for these editions are all quite similar, and it’s hard to tell them apart through the fog of printing and scanning. But the J is noticeably distinct. In SN, AN, and DN, the bowl of the J is like this:


While in MN it’s tighter, curling a little closer to the bottom:


It also seems like the serifs are lighter in the MN font; and the stem is straighter.

Anyway, this confirms my long-held suspicion that the reason for the difference is simply the availability of the letterforms. In the old days, these would be cast as distinct piece of metal, and a print shop would have a limited range of choices. Once the font is fixed, it’s not easy to change it.

Comparing DN vol i and MN vol i, they were published at the same printers, and only two years apart: MN in 1888 and DN in 1890. Yet the typographic styles are quite distinct. DN, for example, uses letter-spacing for emphasis (presumably in the absence of an italic), but this is not used in MN. So it’s likely they were set by different people, and either using their favored fonts, or just what was available.


Thanks Bhante,

So there is no logical reason. Just the available of the type… :cry:

1 Like

But they were once distinct sounds, the ṅ being a velar nasal:

aṅ: [aŋ]
iṅ: [iŋ]
uṅ: [uŋ]

And the ṃ a pure/unmodified nasal:

aṃ: [ã]
iṃ: [ĩ]
uṃ: [ũ]


Sure, but they’re still both nasals, and today are pronounced pretty much the same. The point is that in every other case, “dot underneath” means “retroflex”.