1. We write 'lekha', not 'likha'. But following this pattern, it will be 'likhbo', not 'lekhbo'. The same applies to 'shekha, shikhbo'.
2. The Sanskrit suffix dāniim combining forms idāniim/idāniing, tadāniim/tadāniing; from these come idāniintan and tadāniintan. Therefore, due to the long 'i' in this 'dāniim' suffix, there is no reason for 'idāniing' to become 'idāning'. 'Idāniing' is correct, 'idāning' is wrong.
3. 'Āgāmiikāl' is correct, 'āgāmikāl' is wrong. Being a compound word, it cannot be written separately as 'āgāmii kāl'. Similarly, gatkāl, parbartiiikāl and bigatkāl.
4. Due to Bengali phonetics' natural vowel harmony tendency, the preceding [æ] transforms into [e] under the influence of the following [i]. Thus 'khele' [khyāle], but 'kheli' [kheli]; 'byasta' [byāsto], 'byakta' [byākto], but 'byakti' [bekti], 'byatikram' [be_], 'byatihār' [be_], 'byaṣṭi' [be_], 'byabhichār' [be_]; similarly 'ek' [æk], but 'ekaṭi' [ekṭi].
Now the question is, the poetic non-finite verb 'byāpiyā' should be pronounced [bepiyā] according to this rule, but it's usually pronounced [byāpiyā]. In the song line 'taba nirmal niirab hāsya heri ambar byāpiyā', we indeed hear the [byāpiyā] pronunciation. What about just 'byāpii' (a variant of 'byāpiyā')? ('Byāpii' is pronounced [byāpi].)
Notably, 'byāpiyā' might once have had the [e] pronunciation, as evidenced by the e-vowel in the word 'byepe'. Remember 'bādal-bātās byepe (āmār) hṛday uṭhichhe kẽpe'.
Furthermore, 'byāpti' is undoubtedly pronounced [byāpti], not [bepti]. Then is there naturally a phonetic distinction between ya-phala and ya-phala ā-vowel?
Let's see.
When you've heard 'byakti' as [bekti], and perhaps say it too, then the others could also be heard and said. Many haven't heard 'byathita' as [bethito], but Jagamaya Mitra sings, 'duḥkhaśoke byathita [bethito] chite nāi-bā dile śāntnā'; you can check it on YouTube right now if you want. Where a general tendency of Bengali phonetics is under consideration, personal experience cannot be decisive. And again I say, this tendency exists in verbs and various other words.
5. Unish, untriśh, unchalliśh... up to unnabbi are all tadbhaba forms. (The spelling should have 'u', not 'ū'.) Their corresponding (ordinal) tatsama forms are respectively ūnabiiṃśa, ūnatriṃśa, ūnachatbāriṃśa... up to ūnanbati. (The spelling should have 'ū', not 'u'.) Note the use of 'u' and 'ū'. In tadbhaba words 'u', in tatsama words 'ū'.
6. 'Biṣaphoṛā' versus 'Biṣaphoṛā'
'Goder upar biṣaphoṛā', or 'Goder upar biṣaphoṛā'?
The painful thing that appears on the body is 'phoṛā', not 'phoṛā'. The word 'phoṛā' comes from the root 'phuṛ' (to pierce). 'Phoṛā' is the colloquial form of 'phuṛā'. So there's no reason for biṣa-phoṛā to appear on top of a boil; it should be phoṛā.
Let's look at another interesting thing. From 'bisphoṭak' should have directly come 'bisphoṛā' or 'bisaphoṛā', but 'biṣa' entered and became somewhat meaningful—this is folk etymology; meaning in people's mouths the original word took such a form that people found pleasing or convenient to say and understand.
7. Some say that 'rā' or 're' are not used as vocatives. "Hey rā! Where were you all day?" Again "Hey re! Will you go to watch the game?" — I've heard such vocatives indeed! (Note, not 'baiki', but boiki or boi-ki.) Even (not 'emankii', but emankii) "hā rā, hā re" vocatives I've heard with pronunciation variations.
8. 'Uchit' should be corrected to 'uchit', but bhabiṣyat, jagat, haṭhāt etc. should have 't'. At word endings 'ñja' should be written, because Bengali has no conjunct letter called 'nja'. Words borrowed from English like spanja, bronja etc. have long been written with 'ñja' (though for English pronunciation they should have been written spānja, branja/branja). By the same rule, 'āṅkel' should be written, not 'ānkel'.
9. later - leiṭā(r)
letter- leṭā(r)
latter- lyāṭā(r)
Aṃśata (partially), antata (at least), āinata (legally), āpātata (for the time being), itastata (hesitantly, this way and that), ubhayata (on both sides), kāryata (practically), chaturthata (fourthly), jñānata (knowingly), tṛtiiyata (thirdly), dṛśyata (apparently, visibly), dbitiiyata (secondly), pūrṇata (completely), pratyuta (on the contrary), prathamata (firstly), pradhānata (principally), prasaṅgata (in the context thereof, by the way), phalata (consequently, as a result), baśata (under the influence of), bastuta (in reality), bāhyata (externally), biśeṣata (particularly), mukhyata (chiefly), mūlata (mainly), sambhabata (possibly), sarbata (entirely), spaṣṭata (clearly), sbataḥ (spontaneously), sbabhābata (by nature), sādhāraṇata (generally), rāgata (angrily)
Each of the above words contains the Sanskrit suffix tasil, whose 'il' is elided, leaving 'tas' (taḥ). In modern spelling we omit the final bisarga, otherwise the words would be spelled 'antataḥ' etc.
Among the stems we see, in Bengali usage these can be called pure nouns: aṃśa, anta, āina, kārya, jñāna, dṛśya, bastu, sbabhāba, rāga. With the tasil suffix, all of them (we'll write 'sabkaṭi', not 'sabka'ṭi', because we should avoid apostrophes as much as possible) become adverbs and highly functional words in sentence construction. I've roughly given an English equivalent for each to show the adverbial form.
In Bengali, to create adverbs from nouns often requires the help of adjective + postposition, such as 'aitihasikabhābe' or something similar to mean 'historically', 'tāttbikabhābe', 'tattbagata dik theke' type phrases to mean 'theoretically'. This somewhat impedes the flow of language, affecting conciseness. Perhaps through proper use of the tasil suffix, more effortless and precise words could be created—this is worth considering. The adverb vocabulary is not a closed class but an open class—it can be expanded at will according to desire and need; so this can certainly be thought about.
10. Sarbata can be written separately, though we're accustomed to seeing it in forms like sarbatobhābe, sarbatomukh, sarbatobhadra! "Sarbata mana gala rādhe binodinii rāya/ bṛndābaner baṃśiidhārii ṭhākur kānāi"
11. Let us discuss a rather confusing matter.
Are the -i and -o used after nouns or verbs in sentences . . . (1) suffixes, (2) inflections, (3) postpositive particles, or (4) something else?
Let us see.
While Bengali does have two suffixes called 'i' and 'o', the '-i' and '-o' we are speaking of are not suffixes. Suffixes join with word roots or verb roots to form new words. That is not happening here. The meaning of the noun or verb remains unchanged; only various kinds of emphasis or connotation are added to it.
These two cannot be called inflections either. After words or roots take inflections and sit in sentences as parts of speech, 'i' or 'o' are added as needed. They neither indicate case relations nor tense of verbs.
They are not postpositive particles either. Postpositive particles usually follow nouns or numerical adjectives to indicate the form or nature of the described person or object. But 'i' or 'o' can also follow verbs. For example: You will certainly go there. I haven't seen him, don't know him either.
What then shall we call '-i' or '-o'? My answer: particles. They might be called expressive non-conjunctive particles. '-i' is used for certainty, excess, exclusivity. I will definitely go, no matter what you say, only Ram can do it, etc.
In Bengali, we manage with '-i' the work that Sanskrit accomplishes with the two particles 'hi' and 'eva'. Tvam hi pranah sharire (You are indeed the life in the body), satyameva jayate (truth alone triumphs). On the other hand, '-o' is not merely a conjunctive particle; it is also used in the senses of moreover, even, up to, at all, etc. Even for Shiva it's impossible; you go too; he didn't come even yesterday; I don't know, don't recognize either.
What Sanskrit expresses with 'api', Bengali can express with '-o'. Ramadapi hi tam manye dharmato balavattaram. Even more than Rama... Here '-o' as a non-conjunctive particle expresses this sentiment: 'even' more than Rama, Bharata is stronger in righteousness.
Let me add more. Instead of writing 'Anil-i will go' like this, we write 'Anili will go.' However, there is a slight difference in sei/se-i, tai/ta-i, ei/e-i, etc.
Generally, it is better not to use hyphens (and apostrophes) in writing unless absolutely necessary. 'If your auspicious light shines in the darkness of sorrow / then let that be so' (ta-i) and "Therefore your joy is upon me" (tai). If we can understand the subtle distinction between these two tais, then to explain simply, wherever tahai meaning exists, there should be 'ta-i'; where it means therefore/so, it should be 'tai'. Similarly 'e-i' in the sense of 'ihai', 'ei' in all other cases; 'I wasn't there at that time,' but 'Only he knew the matter.'
Another difficulty arises with 'ye'. The simple wisdom here is that unless absolutely pressed, there's no need to put a hyphen after 'ye'. If it becomes difficult to understand that the word immediately following 'ye' is what is being indicated by that 'ye', then a hyphen can be placed between that word and 'ye'.
Now let me present some lexicographical and other information about our subject -i, -o. Distinguished dictionaries of Bengali, such as those by Jogeshchandra Roy Bidyanidhi, Gyanendramohan Das, Haricharan Bandyopadhyay, Samsad, have acknowledged these as particles. The renowned grammarian Bamandeva Chakrabarti has done the same. Suniti Kumar also placed them among relational or conjunctive particles in Bhashaprakash. 'Chalantika' did not classify them into any category.
There is no disagreement that -i comes from hi. (Why should there be? Ancient Bengali literature contains phrases like 'chalata hi anguli chapi'.) Everyone also acknowledges that -o comes from api, except Haricharan who says it comes from cha. (Let me teach in this gap: tao, or ta-o? 'Didn't you come even then?'—can be written as 'tao ele na?' 'Should that too be shown?'—we'll write, 'ta-o ki dekhiye dite habe?')
Now I come to the aka suffix topic. Teacher, examiner, creator. These are said to have taken this aka suffix after becoming causative. This can be understood by noting the difference between rajaka and ranjaka from the root ranj.
Can the root bal become causative? Monier Williams gives its meaning as to breathe, to live. Can one reach balaka from that meaning? It's not in Kaumudi, not in Paniniyan, only a mention in Bhashaprakash. In Haricharan, there's just a touch in parentheses beside the word krishaka. Can such a big conclusion be drawn based on this little evidence?
Another problem: there already exists a word 'balaka' in the dictionary—the frothing up when boiling milk etc. Then why do we want to bring in a new 'balaka'? If we take ka in the sense of svarthe, balaka would mean emphasis or special force, not emphasizer. But most importantly, when creating a new category, one should consider whether the word fits into any definite category of parts of speech. -ta, -ti didn't fit into any such category, so the category of postpositive particles needed to be created. For words that have been considered particles all along, unless one becomes utterly hostile to particles, there doesn't seem much point in creating new categories for them. If we must create ever-new categories to compensate for rejecting particles, then the question arises whether such rejection can be called wisdom.
That these are particles is now undisputed. In all dictionaries and standard grammar texts, 'i' and 'o' have been called particles. 'i' comes mainly from Sanskrit hi, 'o' from api. What Haricharan said about coming from cha—I think he was then thinking of the conjunctive 'o'. Chakravat parivartante sukhani cha dukkhani cha. (cha: happiness and sorrow.) But our discussion is about non-conjunctive expressive particles. There api should be appropriate. But that's not the main point. In Bengali or Sanskrit, apart from suffixes, inflections, or postpositive particles, all words used in sentences will fall into one of five parts of speech.
English already has three divisions of particles; Bengali has divisions and subdivisions within particles. If any of them fits well, why invent separate definitions unnecessarily? However, to make the description more specific, a sub-name within particles can certainly be created. Balaka is an uncommon word, its etymology is also questionable; using it to explain these two particles would make the matter more complex rather than simpler, I fear.
12. In all these compound words—'khelachhale', 'byatachhele', 'hatachkhara', 'matrachchara', 'byatachchele'—at least one component is non-Sanskrit. Yet in 'khelachhale', 'byatachhele', and 'hatachkhara', the 'chkh' follows exactly the rules of pure Sanskrit sandhi, but in 'matrachchara' this doesn't happen. Similarly, in 'byatachchele' it doesn't happen (and its meaning and usage differ from 'byatachhele').
What could be the phonetically sound explanation for the absence of sandhi in the words 'matrachchara' and 'byatachchele'?
In 'matrachchara', did the conjunct consonant 'tra' become an obstacle? And in 'byatachchele', did the meaning change to become another word, and though compound, it didn't follow sandhi rules—is that why?
13. Professor Ramaprasad Das has paid special attention to the words 'amuk' and 'tamuk' in his book 'Bhashar Baniyad'. These two words exist with considerable dignity in Bengali writing and speech; many other languages don't have such a strong presence of this type of word. In this context, he has raised the question: why shouldn't these find a place in the list of pronouns in Bengali grammar?
This observation seems excellent to me, and the question legitimate. Indeed, we don't see 'amuk', 'tamuk' in the conventional list of pronouns. To this I would like to add the word 'iye'. This too has pronominal usage. The scientist Satyendra Bose apparently had a habit of using 'iye' rather excessively. I had read in someone's writing—Bhaktiprasad Mallick or Purnima Singh, or perhaps Harikrishna Dev—that sentences like 'Take this from iye and give it to iye' were not rare on his lips.
If words aren't meaningful, categorizing them becomes difficult too. 'Iye' can be an adjective: 'The person is somehow iye,' it can be a noun: 'Today what's-his-name, 'iye' came,' and thus can be many things. This is a multipurpose filler word whose existence has long deserved a place in the list from filler to finder.
14. The verb 'halo' should have an o-sound—this is Shankha Ghosh's considered opinion. Despite his precedent, I think it's better to write spelling (especially of verbs) following a principle, however it may look. I'll write 'holो' instead of 'hal' because 'hal' has a different meaning, but I'll write chhilo or dilo, since this creates no confusion. By the same logic, while hab can be written, deb cannot—one must write debo.
There are many examples of this, like writing 'hoy' might suggest a horse, writing 'ay' might suggest income, etc. But it's also true that unless one is utterly foolish, such confusion doesn't arise. Just as in English, hearing "I bow to you" doesn't make anyone think I'm doing something disastrous with bow and arrow. The main point is that the basis of these discussions should be morphology and phonology; 'getting-confused' cannot be the arbiter.
To put it simply: it's not a matter of meaning, but there's a pronunciation issue here. Someone who sees the verb 'hal' and thinks of a plough, whose Bengali pronunciation is 'hal', has a different problem—like English bow (bao) and bow (bo). The pronunciation of the word 'hal' is 'ho'lo', so when confusion arises (which doesn't happen with verbs like 'ay', 'hoy'), making it 'holo' is proper, which is why we change 'deb' (deity) to 'debo'. We write 'hoto' instead of 'hat' (murder) for the same reason.
15. The word is sangopang; there's no alternative to it. Many might absentmindedly write different spellings, and through editorial gaps, these might even survive in print; but I don't think they should be given importance. However, addyaprant and addyopant create the same kind of puzzle.
I don't know if there's any word called 'addyaprant'. This is dvandva compound—does it carry the meaning we want to express, 'from beginning to end'? Sangapang/sangopang is at least a matter of pronunciation, but that's something different. 'Sangopang' also has a quite scholarly usage, such as sangopang Veda.
(To be continued . . . )