 Welcome, I welcome you all to this lecture in the course Sandhi in Paninian grammar. In this lecture, we shall study some strategies of Sandhi's splitting and we shall take some concrete examples. We shall also briefly deal with the scenarios possible situations where Sandhi's splitting is extremely important and needed. In this course, so far we have studied Sandhi in light of the sutras given by Panini. We studied five topics popularly known as Pancha Sandhi, Prakaranam in the traditional Paninian grammatical tradition. We studied At Sandhi, Vavel Sandhi, Hal Sandhi, Consonant Sandhi. Then we also studied Visarga Sandhi and Swadhi Sandhi. Finally, we studied also where Sandhi does not happen, namely Prakriti Bhava, where sounds remain in their own form without getting any modification even when uttered in close proximity or Samhita. These sutras and the Sandhis that are generated by them and the knowledge of them is important for us to know the phenomenon called Sandhi. This knowledge can be applied in order to generate Sanskrit sentences and also in splitting the Sanskrit sentences into its constant events when a particular given text is given or an effort is made to understand a given text. So, here are some strategies of Sandhi splitting and we have taken Srimad Bhagavad Gita as a text to be split. This is the text from which we shall cite examples. First of all, let us study the possible situations where we may face the need of splitting the Sandhi. For example, if one is deciphering a manuscript written in a style which is marked with no full stops at the end of the sentence and the first task for the reader or the researcher in this case is to mark the boundaries of the unit in the form of a sentence. So, splitting the text in sentence units is the primary task for this particular reader and further splitting the sentence unit into words is the next task only then one will be able to make the meaning of the text under study. Here are the screenshots of the two palm leaves and this is a manuscript written on palm leaves. As you see the text is written down horizontally one after another without any visible break in between to indicate a unit called sentence. If you look at these leaves closely you might figure out that there are some marks to indicate a bigger unit in the form of a particular section a particular sutra and the commentary on it. These manuscript pictures reveal also the fact that the manuscripts can get worm eaten. There is this portion over here which is almost broken and illegible and then one has to decipher this particular text from this particular manuscript and in order to do that one need to figure out the boundaries. These two holes that you see in between they are the ones through which the thread is put and then all such palm leaves are bound together. If you notice even the color of the two palm leaves may differ. If you take a closer look at one of them you will find these following pictures and these letters. Now you see these are letters written one after another without there being any mark of any break in terms of a unit. So it becomes a very difficult task for researcher to go through such manuscripts which are the original sources of the texts and then make the meaning of the text. Now in such a scenario Sandhi splitting strategy if known becomes extremely important. Knowledge of Sandhi the sutras etc the examples etc that comes in very handy in this particular task. Here is another clearer picture in which also there are no visible marks indicating the end of a particular unit and so even here we would need information about Sandhi in order to separate units. Here you see numbers 11 and 12. So this is the portion which talks about the 11th sutra in the first sub chapter of the third chapter of the Ashtadhyayi. This is what is written over here and this is what is written over here these are sutras and what follows is the commentary. Now of course in this particular manuscript which is written in a clearer manner one may say that there is some space some additional space that is given at the end of a sentence unit. True sometimes this does help but imagine a condition where those are also not there and then one will be in real trouble and knowledge of Sandhi would become extremely helpful. The second possible situation is that one is reading a printed text. A text is already printed and one is making the meaning of that particular text in order to comprehend the content of it. For example there are several people who want to understand what for example Srimad Bhagavad Gita says. What are the meanings of the words that are part of this particular text? What is the content? The general content many people know about but how does that content come along through the words? This particular intention draws readers and scholars towards such texts. Knowledge of Sandhi becomes the very first tool in their hand and they also sometimes once they gain proficiency they try to interpret the text in a particular way that they want and thereby they learn more about the language and in particular about the text. Thus enriching the overall learning experience and remaining also academically based. This is the possible situation and once again knowledge of Sandhi rules becomes an extremely important tool in this particular endeavor as well. The third possible situation is preparing a computational tool such that if an undissolved string is submitted to this tool as an input it would return the dissolved string as an output. Similarly if separated words are submitted to this particular tool as an input the tool should be such that it would return the combined string as an output. This could be the other possible situation and once again Paninian knowledge of Sandhi rules will come in handy in this endeavor as well. Now let us take examples from Srimad Bhagavad Gita and try to figure out strategies with the help of which Sandhis can be identified and split. These strategies could be helpful in all the three scenarios that we described earlier. Let us take the very first line. This letter Mo and S they are put in a different color blue color to indicate that they could be the possible candidates of Sandhi or a substitute. So, what is the mark? How do we identify that there is a Sandhi over here? One simple clue that is provided in the printed edition in this case is that there is an S in between two letters. This generally would not happen when S is put in between two letters that is a primary indication that there is a Sandhi over here. This is further confirmed when observed that the previous vowel is O in Prathamo. Now we know that this vowel O can be dissolved as one substitute in place of two substitute elements. This kind of scenario. So, O will be a combination of O plus O and however we should note that this can happen in a specific environment namely that O should be at the end of a Prathamo or should be at the beginning of the next Prathamo. In this particular environment O can be said to be combined of O plus O. So, O is the substitute that comes in place of two substituents O plus O. O is at the end of a Prathamo and O is at the beginning of the second Prathamo. If we apply this knowledge to Prathamo Dhyayaha, we note that Prathamo Dhyayaha can be split into these two components Prathamo and Adhyayaha. Now if we check if these two could be possible paddhas, we come to know that Adhyayaha is one slash one of Adhyaya. Adhyaya means a chapter as given in the lexicon. So, as far as the right hand side is concerned its form, its shape is confirmed. Now if we go to the left hand side Prathamo, we note that Prathamo is not part of any derived word. However Prathamo has got a subword string Pratham which matches with the part of the lexeme Prathamo meaning first. We note this and then we move ahead and concentrate on this O in Prathamo once again. Now this O coming at the end of a paddha is more likely to be a sandhi rather than a subanta form of Prathamo. Why we say so is because there is no such form which is a subanta that is derived from Prathamo that is found in the corpus and also lexicon. Given this fact the probability that this O is a substitution in place of two substituents namely O and O takes us forward to split Prathamo into Prathamo plus O. Now if we put Prathamo plus O in place of Prathamo because we split Prathamo Adhyayaha as Prathamo and Adhyayaha. Now this Prathamo can be replaced by Prathamo plus O and so now we have a scenario where we have O plus O at the end of the paddha followed by another paddha that begins with O. I repeat O plus O at the end of this first paddha followed immediately by another paddha which begins with O. So we have O plus O plus O this environment. Now we know that in this environment O at this particular position at the end of the paddha and in this environment meaning thereby that preceded by short O and followed by another short O. We know that this environment is the output of the sutra 61113 So now we go back to the input of 61113 which is RU in this environment and at the word final position and so we get O plus O plus O as the substitute or the sandhi and we go back to the input and that input is O plus RU plus O, O plus RU plus O. Ator or Apuluta the Apulute says that this RU coming at the end of a paddha which is preceded by short O and followed by short O is substituted by O. So this is the output. Now we are going backwards from the output we are now going to the input while learning the sutras our direction was from the input to the output. Now we are going from output to the input. After having come here after having reached this particular stage we further go back to another sutra namely 8266 whose output matches with this particular string O plus RU plus O and we know that the input for 8266 is SIR for the output RU. So O plus RU plus O is the output of the input O plus SIR plus O. This SIR occurring at the end of the paddha preceded by short O and followed by another short O. We have a clue here to recognize that 8266 applies over here. We discuss this fact when we studied this particular sandhi we said that this is an exception to the asiddha principle. 8266 applies first and produces RU as the output. Now this RU becomes an input for 6113 which provides the output in the form of O. This is the exception to the asiddha principle. Now we also know that SIR is a 1 slash 1 suffix as per S412 and then it is SU stated over there with O as a marker marked by 132 or 9. So now we are sure that this is 1 slash 1 of prathama which is otherwise shown as prathamaha. We also find that this 1 slash 1 in prathamaha matches with 1 slash 1 of adhyayaha. This matches with the unwritten rule in Sanskrit that qualifier and qualified appear in the same case where prathamaha is the qualifier and adhyayaha is the qualified. So the fact that 1 slash 1 exists in prathamaha as well as adhyayaha this fact matches with this unwritten rule. Then we look at the meaning and find that the procedure that we adopted is producing a compatible meaning combination namely now begins the first chapter. Atha means now begins and prathamaha means first and adhyayaha means chapter. Now we also check with the context and find that this meaning fits in very well or rather perfectly at this place as this is the beginning indeed of the first chapter and this line is a declaration of this particular fact that now the first chapter begins. So to sum up we can take prathamodhyayaha as an input and then we can split the sandhi in the following manner. So prathamodhyayaha is first split into two prathamo and adhyayaha and prathamo is put into square brackets adhyayaha is also put into two square brackets and there are two more square brackets at the end they indicate that this is one unit one sentence. So now we have prathamo prathamo adhyayaha this prathamo is further split into prathamo plus u plus adhyayaha now this u is a substitution in place of ru so the next step would be prathamo plus ru plus adhyayaha and this ru is a substitution in place of sir so we say prathamo plus plus adhyayaha. So prathamo plus sir plus adhyayaha generates ru as the substitute in accordance with 8266 then this ru gets substituted by u by 61113 then this u and this ur they get substituted by o by 6187 and finally this o and this ur they get substituted by this o by 6109 so finally we arrive at the two constituents written down as separate padas and they are prathamaha and adhyayaha prathamaha and adhyayaha this is how we split at the prathamo adhyayaha we also check the corpus when we apply this test of corpus we find that there are similar sentences at the beginning of each chapter like adhatrutiyodhyayaha adhachaturthodhyayaha adhapanchamodhyayaha and so on and so forth the position of these sentences and the commonality in wording allows us to figure out the different word in those sentences and the meaning with the help of the dictionary trutiyayaha means third chaturthaha means fourth panchamaha means fifth etc this is how the sandhi gets split with the help of these strategies in this particular example let us now take another example the next sentence in shreemad Bhagavad Gita dhritarashtra uvacha so what could be the possible analysis of both these words dhritarashtra can have at least two explanations the first one is that dhritarashtra is the vocative singular of the word dhritarashtra and the second explanation is that this is in most parts similar to one slash one of the word dhritarashtra if we look at uvacha this also has got two explanations first one is this uvacha is third person singular karta of perfect past of the verbal root meaning to speak this word expects the nominal root to specify who this karta of the action of speaking is and this karta is also expected to be in one slash one the second explanation of uvacha is the one uvacha is first person singular of karta one p slash one karta of perfect past of the verbal root uvacha to speak this word expects the nominal root to specify who this karta of the action of speaking is this karta is also expected to be in one slash one but in this case the choices are clear because this first person singular is nothing but i so the meaning of this uvacha is i spoke without uttering the word i so now we have four options two options for dhritarashtra and two options for uvacha so there are four possible combinations noted down on this slide as a b c b a consists of the first explanation of dhritarashtra and the first explanation of uvacha b consists of the first explanation of dhritarashtra and second explanation of uvacha c consists of the second explanation of dhritarashtra and the first explanation of uvacha and d consists of the second explanation of dhritarashtra and the second explanation of uvacha these are the four possible explanations let us see which one of them seems more probable amongst these four only c seems more probable and also compatible and also most fitting in the context where the next where in the next verse it is understood that someone is asking somebody else a question so obviously a b and d get eliminated on account of them being less probable less compatible and not fitting in the context now we look for a sanny so one slash one suffix is stated to be sir but it is missing here so we conclude that this is deleted which means that this is a zero substitution now we already know that if we have this particular scenario a plus zero at the end of the paddha plus u at the beginning of the next paddha if this is the environment then this is the output of a319 lopashakalyasya whose input is a plus here plus u a plus here at the end of the paddha so now from dhritarashtra uvacha we go backwards and generate a plus here plus u we also know that a plus here plus u is an output of a317 whose input is a plus ru plus u so now we go to the next stage a plus ru plus u dhritarashtra plus ru plus u we also know that this is an output of a266 whose input is a plus sir plus u and this sir is one slash one suffix that gives us the word dhritarashtra plus sir which is nothing but dhritarashtraha if written separately so we can show this process of splitting in this particular manner in this sequence dhritarashtra uvacha is the string and we note that there is zero substitution this zero substitution is the output of the input dhritarashtra plus here plus uvacha because a319 applies then dhritarashtra plus here plus uvacha is the output of the input dhritarashtra plus ru plus uvacha in accordance with a317 and we also know that this dhritarashtra plus ru is the output of dhritarashtra plus sir plus uvacha in accordance with a266 now here there is no question of asiddhatva the rules are followed in order they apply in order so now after having split this we know now that the two separated words are dhritarashtraha and uvacha this conclusion confirms that it's fit it fits perfectly in the context as now we know that it is dhritarashtra who is third person singular who is the karta of the action of speaking that happened in the past perfect so the meaning of the sentence is dhritarashtra spoke also the context supports this explanation as this is the name of the king who in the next string namely dharmakshetra kurukshetra samaveta yuyutsavaha mamaka pandavaschaiva kim akurvatasanjaya is asking this sanjaya about his own sons and those of his brother so we confirm this particular split as finally accepted we also apply the test of corpus and if we check the corpus we find similar sentences at the beginning of each chapter as well as in the middle suggesting the speaker change or name of the new speaker etc for example sanjaya uvacha or arjuna uvacha all such instances can be explained in the similar fashion sanjaya uvacha sanjaya 0 uvacha sanjaya ya uvacha sanjaya ru uvacha sanjaya sa uvacha sanjaya ha and uvacha similarly arjuna uvacha also the position of these sentences and the commonality in wording allows us to figure out the different word and its meaning with the help of the dictionary sanjaya for example arjuna for another example we also observe that there is a sentence shree bhagavanu vacha however there is no sandhi over here this is how we split sandhi in cases of this kind to summarize application of rules in reverse order than what we have studied so far may help in splitting the sandhi in accordance with the grammar of panini we also observe that other resources like the dictionary of prathipadikas as well as that of subanthas is also very helpful in this task to eliminate options based on probability measures the split that we arrived at also stands the test of the context very well one important point to remember over here is that the multiple possible splits form the basis of different interpretations by different scholars that are available in the tradition thank you for your attention