Crash course on reading Sanskrit Heritage Corpus

To whom is this course adressed ?

This course is adressed to an English-speaking beginner. There is no prerequiste. In particular, it does not require familiarity with the devanāgarī writing system, which will be progressively acquired. We only assume some acquaintance with basic notions of linguistics.

What is Sanskrit ?

Sanskrit was the learned language of Ancient India, frozen twenty-five centuries ago by a genious linguist named Pāṇini. Pāṇini managed to model a high-register form of the vernacular language of his time as a functional generative grammar. This allows to derive formally a correct enunciation of the language as a structure expressing the communication intent of a speaker. From this structure may be extracted both the phonetic utterance and its meaning, expressed as a dependency graph between participants. This grammar, refined by a lineage of further linguists, is precise enough to serve as a blueprint of computational linguistics software, such as the tools which will be demonstrated in this session.

Methodology

We are going to read a text with the help of the Sanskrit Heritage software. In doing so, we will learn the basic sentence structure of Sanskrit, we will recognize the morphological structure of the words of the vocabulary, and we will be able to consult their meaning in the dictionary "Sanskrit Heritage", which will also provide the semiotic indications allowing to understand the text in the Indian tradition. It is also possible to look-up meanings in the English Monier-Williams dictionary, although this will reduce the functionalities.

The course uses hypertext technology. The page you are reading in this moment with your Internet browser gives the instructions on reading the text and its interpretation. By clicking on its links other windows will appear, allowing you to access the software features.

First reading : Vikramacarita

Our text is taken from a work called Vikramacarita, which has no precise author or date. It's a collection of tales about an ancient king, Vikramāditya, the Sun King of Indian tradition, actually more the myth of the magnanimous king than a precise historical figure.

Some indications on this work can be consulted by clicking on the link in the previous paragraph. But don't get lost in the forest of hypertext links in the dictionary, close its window, and come back here for our read. This text, taken from the 24th section of the Vikramacarita, follows the version of the Sanskrit Chrestomathie by Nadine Stchoupak, published in the collection of the Institute of Indian Civilization published by the Librairie d'Amérique and Orient Adrien Maisonneuve, Paris, 1977.

We will first read the text analyzed by our tools, sentence after sentence. To do this, click on the following link to open a new window (right click). How Śālivāhana solved the mystery of the strange inheritance.

You have in front of you in the new window the list of the 39 sentences which constitute our extract. They are presented in romanized characters according to the IAST standard. Each Sanskrit phoneme is noted by a Roman character possibly provided with diacritics. It allows you to easily read the Sanskrit, which is written as it is pronounced. This is an obvious advantage over English, whose prononciation has diverged over time and location. On the other hand, English writing presents the utterance as segmented into individual words, and liaison is implicit, whereas Sanskrit records the unsegmented phonetic utterance, and thus word boundaries have to be guessed. This is where our software will help you.

The protocol to follow for each of the sentences is as follows: first, read aloud the sentence as it is written in romanized notation, and repeat it until it becomes fluid. Then click on the link. Keep in parallel the two windows of your browser, that of the course explaining the content of the other, which contains the Sanskrit text, with hypertext links giving the grammatical information.

Now click on sentence 1. The page that appears first gives the sentence in writing devanāgarī (blue). In a second phase of the course, we will learn to read this writing, which is syllabic writing. Each character represents on the one hand the sequence of consonants of the syllable by a ligature, and on the other hand its vowel with a diacritical mark. The initial vowels have their own sign.

Ignore the "Unique Solution" and "UoH Analysis Mode" indications for now. Then appears the analysis of the sentence, cut into a series of words. The words appear in rectangular colored boxes, aligned with the sentence written in romanization. We see in the analysis of the first sentence 4 colors of boxes. Red boxes contain verbal forms. Purple boxes are language tools, like adverbs, prepositions, conjunctions. Blue boxes contain nouns or adjectives. Yellow ones contain word themes to form compound words. A compound word is represented by a yellow box followed by a blue box. So here the sentence possesses 7 segments, each in its box, but only 6 words, the two segments purandara and purī forming the compound word "purandara-purī". Notice that they are glued together in the devanāgarī writing of the sentence. This indication is important for the correctness of prosody. The fragment "purandarapurī" is a non-breaking chunk of writing, a compound word not admitting a pause between its segments. However, you can resume your breath between the segment "purandara" and the segment "purī" on which you put the accent, because here the meaning of the compound is a specialization from the word "purī", city, by its name Purandara. Practice speaking aloud.

By clicking on a box, we get the lemma of the word, giving its stem as a link in the lexicon, and the variations allowing the production of the specific form of the word. So the first word is indicated as the base form "vikramāditya" in the genitive masculine singular. It therefore indicates the possessor of the meaning of the following word, which is rājye, locative form of the stem "rājya" (kingdom) in the locative. The first two words therefore form the adverbial clause "in the kingdom of Vikramāditya". By clicking on the link of the "vikramāditya" stem you access the corresponding entry dictionary which gives different indications on the king Vikramāditya.

When you access the dictionary, you can go back to the analyzed sentence by the previous page link of your browser. The pop-up window giving the lemma of a word can be closed by clicking on the cross sign ✘ or by clicking on another box.

Let us continue reading. The next word is the compound "purandara-purī", which denotes the city of Purandarapurī, capital city of King Vikramāditya in Mālava, also called Ujjayinī, etymologically "Victory", an agent noun of verb "ut-ji" (to win). Ujjayinī is modern Ujjain, where one can still see, in a pavilion near the sacred lake, a figuration of Vikramāditya and the presumed remains of his throne.

The original throne, in the golden legend of Vikramacarita, was carried by 32 magical statues. When the king sat there, he would magically become infinitely magnanimous. Centuries later when the king Bhoja wanted to sit on it, each of the statues in turn came alive and told a legend about Vikramāditya's greatness of soul. The set of 32 stories constitutes the Vikramacarita, of which we read an excerpt from the 24th.

The following word, "nāma", mauve, is a clitic particle meaning "by name", and qualifying the following word nagarī, which means "city" or "capital". Note that the name of the writing "devanagarī" is a compound "deva-nagarī" meaning "(writing) of the divine city". The sequence of words "purandarapurī nāma nagarī" thus forms the noun phrase "the city named Purandarapurī", and the nominative case of "nagarī" indicates that it is the subject of the sentence. The sentence ends with the red verbal form samabhūt, which is a past tense form of the verbal root "bhū_1" (to be), preceded by the particle "sam".

Click on the red box to reveal its lemma. You can access the verb "saṃbhū" in two clicks from there: by clicking on "bhū_1" you get the entry from the root "bhū_1" in the lexicon. If you have chosen the Heritage dictionary, this entry carries the list of all prepositions that can be used with this verbal root: (ati, anu, abhi, ā, āvis, ut, tiras, parā, pari, pra, bahis, vi, sam). The last one is "sam", and by clicking on it one accesses the verb "saṃbhū", to exist. If you have instead chosen the Monier-Williams dictionary, this kind of navigation is not available, and one has to search the entry "saṃbhū" using the Index search. By two backtracking of your browser you will return to the analysis page of the sentence.

In order to understand morphological lemmas, it is necessary to become familiar with a certain number of abbreviations of a linguistic nature. The list of (French) abbreviations of the system is available as a pdf document. We see for example that "aor." means "aoriste", i.e. English "aorist". Don't worry about the terminology of this past tense, called aorist by linguists, by analogy with the nomenclature of ancient Greek. The aorist is not a very common tense in classical Sanskrit, we will see little of it in this course. The usual times of the past are the imperfect, a general preterite, often imperfective, and the perfect, time of a bygone past, with a perfective aspect.

The indication "[1]" following "aor." indicates the formation paradigm of the aorist, of which there are 7 varieties. It may be ignored. The rest of the lemma of samabhūt tells us that the form is in the active voice, in the 3rd person singular. Which is consistent with the lemma of its subject nagarī with nominative singular. Actually, the agent of the action is indicated by the subject in the nominative, and it must agree in number and in person with the verbal form. There are 3 genders in Sanskrit, masculine, feminine and neuter. Gender in Sanskrit is arbitrary, except for the referents of persons or animals, where sex determines gender. Neutral is the kind of abstract nouns.

We have used the terminology "nominative" which is familiar to all who have notions of Latin. For students unfamiliar with the concept of declension, it suffices to know that the substantive words are classified in 7 cases, called nominative, accusative, instrumental, dative, ablative, genitive, locative and vocative. The case is used to designate the syntactic function of the word. For example, the nominative is used to designate the subject of the verb, and the accusative its direct object complement. In French or English, these roles are implicit from their position in the sentence, while in Sanskrit they are explicitly encoded in the word, as a suffix characteristic of the case. The advantage is that the word order is free to a great extent. This allows to reveal the information provided by the sentence more flexibly, to do stylistic effects and conforming to the meter for rhythmic poetry. We will gradually come back to these grammatical notions, let's close this linguistic parenthesis and get back to the text.

So now we get the meaning of the entire sentence: "In the kingdom of Vikramāditya was the city of Puraṃdara, its capital". We can now continue reading, by clicking on the button marked "Continue reading" in the sentence analysis page, which brings us back to the summary page of the text, where you can now go to the next sentence.

Don't forget to say aloud the sentence you just finished to understand, by adapting the prosody to the syntactic structure. Then read aloud the new sentence.

In the second sentence we recognize the colors of the first. Similarly to it, its verb is at the end of the sentence, here asti, form of the frequent verb "as_1", to be, in the 3rd person singular: (it) is. Its subject is the previous word vanik, merchant. It is himself preceded by kaścit which is an indefinite qualifier "some", playing the role of our indefinite article "a/an". The previous word is a compound adjective mahā-dhanikaḥ, formed of dhanikaḥ, "rich", with nominative singular, preceded by the bare stem mahā of the adjective "mahat" (big). The meaning is therefore "very rich".

The first word of the sentence, tatra, is the adverb of location "there" which connects to the preceding sentence. So we get the meaning "There lives a very rich merchant". This right to left analysis of the sentence is the standard method. We look for the verb at the end, easy to recognize since it is red, and we go up the sentence in search of its actants. So far we have had intransitive verbs, which have only one actant, their subject, in the nominative case. Transitive verbs have two actants, the agent of the action and its goal, or direct object complement, which is declined in the accusative in active voice. But, in the passive, which is in fact much more frequent, the subject is the goal, but the agent is in the instrumental, as we will see shortly.

Note that the written form of this sentence has only three chunks. The last one indeed coalesces the three words kaścit, vanik and asti, by performing a phonetic smoothing: the final mute "t" of "kaścit" becomes the corresponding voiced consonant "d", on contact with the initial semi-vowel "v" of "vanik". This phonetic smoothing, called "sandhi" (junction), is the generalization of the inter-word liaison. We can see that writing is taken seriously as a faithful representation of oral utterance. Hence the ease of pronouncing this written form without ambiguity. On the other hand, we pay for this advantage by the difficulty of decipher this oral enunciation as a series of words. This difficulty is combined for non-Indians with the deciphering of the Devanāgarī script, specially when glueing two words, the first of which ending with a consonant, and the second starting with a consonant too, because the two syllables in contact collapse into one, and the resulting ligature conceals the border of the two words. Fortunately, our software overcomes the difficulty, by automatically performing decryption. You will gradually learn to discern on your own these glueings, and to undo them. For this, it is essential to pronounce aloud, first in word-by-word fashion, then in continuous speech.

It is time to clarify the notion of a word, which is ambiguous in English. It is essential to distinguish the inflected word, by declension or conjugation, as it appears in the sentence, from its stem which is referenced in dictionaries. In the following, we will use the Sanskrit "pada" (step) to designate the inflected word. Our software performs the "padapāṭha", etymologically the step-by-step path of the enunciation, by undoing the sandhi links between his constituent padas.

Let's continue reading with the 3rd sentence. A new color appears, sky blue, which indicates a pronoun, here "tad" (personal "he/she/it" or demonstrative "this") declined in the genitive tasya (of this). The verb is the same as in the previous sentence, "as_1" (to be) but now plural " santi " (are). Dictionary lookup tells us that "catur" means number "four", and "putra" means "son", hence the nominal phrase catvāraḥ putrāḥ, "four sons", subject of the sentence, which therefore means "He has four sons".

In the following sentence, we find this same nominal phrase , but now in the accusative case: caturaḥ putrān. It is followed by the absolutive form āhūya from the verb "āhū" (to summon), of which it forms the object complement: "having summoned the four sons". This absolutive clause shares its agent with the main verb of the sentence: avādīt (he says). It is preceded by the temporal adverb maraṇa-samaye, compound word in the locative expressing "on the verge of death". The subject of the sentence is the nominal phrase made up of 4 words in the nominative: vṛddhaḥ sa vaṇik vyādhitaḥ (old, this merchant fell ill). At the beginning of the sentence, we find the adverb tataḥ, which indicates the temporal sequence (then), followed by a nominal group in the locative mahati kāle gacchati (a long time having passed). We therefore translate the sentence with "Many years passed, and the old merchant fell ill; at the time of his death, having summoned his four sons, he said to them:"

This sentence has the following characteristic structure of story telling. First of all, we situate the action by a series of clauses in the absolute locative. Next, a series of absolutive clauses describes the sequence of actions in the story, terminated by the main clause which alone carries a verb declined in the person of the agent: "having done action 1, ..., having done action N, he did action". In the English translation, we replace the sequence of absolutives by the sequence of actions, with forms conjugated in the time of the main clause: "he did action 1, ..., he did action N, he did the main action".

Two difficulties are to be noted. The first is the ambiguity of the pada kāle, visible by clicking on its blue box. Indeed it may be a form of the substantive "kāla_1" (time), or a form of the adjective "kāla_2" (black). The two meanings being very common, only the context allows them to be disambiguated. The second is that "gacchati" is a very common form (but irregular) of the present tense of the root "gam" (to go): "he goes". Here, this interpretation was not retained, because the verb "gam", transitive, asks for an additional object to indicate where the subject is going. There is no nearby noun in the accusative to fulfill this function, but the locative form of this present participle (going) agrees with the phrase "mahati kāle" (long time) to mean "passed". Here the ambiguity is not apparent, because we have chosen to put gacchati in a blue box, as a nominal form, rather than in a red box that would suit the verbal form "he goes". We see here that we have simplified your task, but that when you are faced with an un-analyzed text, all kinds of ambiguities are likely to lead you astray. Our method allows to acquire first the understanding of the sentence, partially disambiguated, in order to be familiar with the syntactic turns of Sanskrit, before leaving you faced with the combinatorial complexity of raw sentences that are non-analyzed and even non-segmented in individual words.

Let's move on to sentence 5. Remember to always say it out loud. We first see a new color, with two green boxes which indicate, first an interjection bhoḥ, which corresponds to our "hey", then a vocative putrāḥ, in the same form as the nominative plural of the word "putra" already encountered. Here we must expect an interjection of the father addressing his children, such as the avādīt from the previous sentence. We leave the impersonal style of the story to switch to dialogue mode: the father is now the speaker of the sentence, and he addresses his children with "Hey, (my) sons". Then we find the personal pronoun mayi of the first person (the speaker), in the singular locative. It qualifies the next word, mṛte, locative form of "mṛta" (dead), and we recognize our friend the absolute locative, giving the context "at my death" for the rest of the sentence.

This "mṛta" is a past participle of the root "mṛ", to die, like our adjective "dead" as a matter of fact. This "at my death" is very exactly "me dead". Here you must be wondering how to pronounce this letter "ṛ". Well, this is a vowel, vowel form of the sound "r" which must be pronounced rolled. Actually, at the strong vowel degree (like "e" is for "i" and "o" is for "u"), we find "ar", pronounced like a silent "e" followed by a rolled "r". This therefore suggests pronouncing "mṛta" like the beginning of "murder". This verbal root goes back to the oldest branches of the tree of Indo-European languages, and we find it in Latin mortus, which gives muerta in Spanish and mort in French. These etymological considerations are important, because they help to acquire the vocabulary of a language from the same linguistic family as one's mother tongue.

In fact this vowel "ṛ" is the phoneme of Sanskrit with the most variations according to the speaker. The name of the divine hero Kṛṣṇa is pronounced in India Krishna or Kroushna depending on the region. Let's take the opportunity to visualize the "great incantation" mahāmantra: hare k.r.s.na hare k.r.s.na k.r.s.na k.r.s.na hare hare | hare raama hare raama raama raama hare hare ||. Surprise: all boxes are green, this is an invocation of Viṣṇu alias Hari, of which Kṛṣṇa and Rāma are two incarnations (avatāra). We notice in passing that "hare" is the vocative of "hari".

After this interlude, let us close the window of the mantra that has just opened, and let us come back to sentence 5. The two blue padas in the genitive plural, recognizable by their final "ām", that is to say "bhavatām caturṇām", mean "of you four", the word "bhavat_2" being the polite form of "you" in the third person, similar to English "Sir". These genitives qualify avasthānam, form of the neutral word "avasthāna" which can be translated by social situation or goods. The mauve pada ekatra is a locative adverb built on "eka" (one), meaning "in one place". For "bhavati" which follows, as above we have the choice between the personal form "is" or the present participle "being". In the first choice, we would have two sentences (red "bhavati" ending the first sentence); in the second choice, which is the one in our analysis, we have a participial clause subordinate to the main clause: "the estate of the four of you remaining undivided". The main clause has for verb the last pada, red bhaviṣyati, which is in the future tense of the verb "bhū_1" of which we have already seen several forms. Let us note in passing that "bhū_1" is the verb "to be" in the dynamic sense of becoming, while "as_1" that we saw above with asti and santi is the verb "to be" in the static sense of existing. The subject of "bhaviṣyati" precedes it, it is vādaḥ, nominative singular of "vāda", which means "debate, dispute". The rest of the sentence is a sequence of grammatical words. The conjunction vā marks disjunction, the particle "na" is negation, the sequence vā na vā means "yes or no", which is generally translated by "maybe". The particle hi is our "really", and the adverb "paścā" corresponds to "later".

We now understand the sentence : "Dear sons, on my death, if your estate remains undivided, later there may ensue disputes".

Let's move on to sentence 6. The verb is asmi, the form of "as_1" at the first person. Its subject is kṛtavān, a past active participle of the very common root "kṛ_1" (to do). The phrase "kṛtavān asmi" (I am having done) is a somewhat solemn way of saying "I am the one who made", that we will replace in English by "I carried out". What the merchant did, the object or purpose of the action, is given to the previous pada vibhāgam, form in the accusative of the word "vibhāga", the share. The above pada is the compound jyeṣṭha-anukramam, where "jyeṣṭha" is the elder and "anukrama" is an enumeration. It agrees with sharing, and the full nominal group therefore means "sharing by order of seniority", the aim of the father's action. The two preceding padas have already been encountered in the previous sentence, they mean "of you four", and in good English we will translate by "sharing between you four". The initial clause is formed, as in the previous examples, by the present participle "jīvat" in the nominative jīvan (alive), preceded by its subject aham, first person pronoun, "I". which gives "me alive". The negation particle na is here combined with the particle iva ("actually") to mean "barely", "just".

This sentence is linked to the previous one by the discourse connective tarhi , which means "therefore". We can now translate the entire sentence: "That's why, still alive, I effected the sharing between you four by seniority rank". We can clearly see in this example the advantage of reading Sanskrit "backwards", in order to understand the predicate (the verb with its possible direct object) before its subject. In translation, we re-establish the sacrosanct subject-verb-object English order. But Sanskrit does not strictly obey the order subject-object-verb that we have seen so far. Indeed, the case variations provide for the semantic role of the concepts brought into play in the sentence, and its syntax is relatively free. The important thing is to see the predicate, verb or participle, and then recognize its arguments and complements. Our software does the work for us, puts in evidence the action expressed by the verb with its blaring red color, which governs the rest of the actants, colored blue, and circumstantial adverbs and other conjunctions, colored purple. The vocatives are distinguished in green, because strictly speaking they are not part of the sentence, they are interjections not participating in its logical structure.

Until now, we have carried out the morphological analysis of padas from verbal roots or nominal or pronominal stems. So we saw that asti, asmi, santi were conjugated forms of the root "as_1" (to be). The software also allows us to generate mechanically all these forms starting from their roots. For example, let's click on the red box asmi of the current sentence, to find out its lemma: [as_1] {pr. [2] ac. sg. 1}. The root "as_1" is a link on the corresponding entry in the lexicon, where we read: "√ अस् as_1 v. [2] pr. (asti)". The number 2 indicates the conjugation class in the present tense of this root. It is itself a hypertext link taht returns to you all the conjugated forms of the root. Click on it. The first array, giving forms of the present tense in the active voice, gives us the forms "asti", "asmi" and "santi" in the expected locations. We notice between the singular and the plural the existence of specific forms for the subject in the dual, when there are exactly two agents. We will ignore these rare forms in this first lesson.

Now go back in your browser page to return to the page of the current corpus, and let's move on to sentence 7. We are a little taken aback, because we look in vain for the red box, this sentence does not contain a verb. Indeed, it is a nominal predication, the predicate being here the past participle of the verb "ni-kṣip". By clicking on "kṣip" in its lemma, we find its definition as "lancer" in Heritage, or its English equivalent "to toss", in Monier-Williams. But in Heritage, you have an extra information in the list "pf. (adhi, ava, ā, ut, ni, pra, prati, vi, sam)", which indicates all prepositions that can be used with this root, as a list of links. Clicking on "ni" we arrive at the special meaning of "nikṣip", "to deposit". The form nikṣiptāḥ is therefore the predicate "deposited". And its subject precedes it, it is the nominal group catvāraḥ bhāgāḥ, that is, four parts. A locative clause indicates where these four shares have been deposited, adhaḥ, meaning "under". Under what? As usual, under what precedes, "adhaḥ" is a clitic preposition, or rather post-position, in the general spirit of giving a notion preceded by its argument. What precedes is here the nominal phrase caturṇām pādānām, which designates a four-feet item, that is to say a bed. What bed? The one indicated by the previous word in the genitive mañcakasya. Its stem is "mañcaka", which here denotes the terrace of the house. Everything is preceded by ofthe locative adverb atra (here, thus). We put everything in its place, and we understand: "Thus under the bed of the terrace four parts were deposited." But often it is more natural to put the passive phrase back to the active voice, with a verb conjugated in the past tense replacing the participle: "Thus I placed the four parts under the bed on the terrace."

Sentence 8 is not a problem. The verb gṛhṇīdhvam is an imperative form of the root "grah", to take (which you recognize as related to the English "to grab"). The noun is a compound word "jyeṣṭha-kaniṣṭha-krama" in the instrumental giving the adverb "by order of elder to younger", similar to "jyeṣṭha-anukramam" already met. Which gives: "Take (your share) by order of seniority."

Sentence 9 bears a new color, orange. The morphological process involved is the formation of compound verbs, from auxiliary verbs "kṛ_1", "as_1" and "bhū_1", by prefixing them with a special form of a noun, usually terminated by "ī", to denote the verbs to do, to be or to become the notion denoted by this substantive. Linguists talk of inchoative verbs. A typical example bhasmīkṛ, to reduce to cinders, constructed from substantive bhasman (cinders) with auxiliary verb "kṛ_1". In our text the noun is "aṅga_1", which means portion. The compound verb is therefore "aṅgīkṛ", to share in parts, and aṅgīkritam is its past passive participle, meaning "sharing done". The pronoun taiḥ ("by these") refers to the choices made by the sons. Using an active twist, one obtains: "Thus, (your) choices will have carried out the sharing".

Sentence 10 is again a nominal passive sentence, a very common turn of speech in Sanskrit. The past participle "sthita" from the root "sthā_1" (to remain, to stay) in the nominative plural sthitāḥ has for subject the group catvāraḥ bhrātaraḥ of the four brothers. We note in passing the resemblance between the word "bhrātṛ" (in the form "bhrātā") and English "brother". Like the daughter "duhitā", etymologically "she who milks (the family cow)" is very close of "daughter".

The story continues in the third person, with tataḥ (then) which connects with the previous episode by an absolutive clause giving the new circumstances: "him having gone to the afterlife". The main clause can be renderes as "the four brothers let a month pass". Obtaining: "When he died, the four brothers let a month pass."

The next sentence literally means "Then, of their wives mutually the quarrel arose", and in better English, active voice: "Then their wives quarreled with each other." The form teṣām of the personal pronoun "their" must evoke in you, by its final "-ām", the genitive plural already encountered with bhavatām, caturṇām, pādānām, and now strīṇām. But the pronouns (as well as the substantive feminine "strī" (woman), have particular frozen forms that are needed to be learnt by heart, because they will be frequently encountered. It is therefore time to look at the entry of the 3rd person personal pronoun in the dictionary and study its neuter ("tad"), masculine ("sa") and feminine ("sā") forms by clicking on the corresponding gender, respectively on the red links n., m. and f.

Sentence 12 is an interrogative, preceded by the interrogative pronoun kim ("what", in the neutral; the masculine stem is "ka", and the feminine one "kā"). Here it translates to "why". The verb is "kṛ_1" (to do), already encountered, but now conjugated in the passive voice, and having for subject kolāhalaḥ, an onomatopoeia meaning a din, in the nominative. We therefore have an exhortation: "Why is there so much noise here?" which expresses the irritation of the husbands.

Sentence 13 is built on the now well-known model: a clause in the absolute locative built on a present participle, followed by a passive proposition built on a past participle. The particle eva, obtained by welding the preposition "ā-" (expressing a movement towards the locutor) and the particle iva already encountered, is an intensifier signifying "even, again, precisely". The pronoun form asmat is the ablative of the personal pronoun of 1st person ("aham" in the nominative singular, "I"), meaning "of us". It qualifies the sharing "vibhāga". The nominal phrase pitrā jīvatā eva in the instrumental ("father being still alive") is the agent of the principal clause in the passive voice vibhāgaḥ kṛtaḥ ("sharing done"). The adverb of time pūrvam indicates the past ("previously"). We can now reorder the components of the sentence in an active turn, like: "Formerly, our father being still of this world, he effected the sharing between us four." In this sentence, we remark again the similarity between "pitṛ", father, and the cognates Latin "pater", French "père", English "father", German Vater.

It will also be noted that the segments jīvatā and eva are glued by sandhi in the fragment "jīvataiva" from the original text. More complex is the form "jīvanneva" obtained in sentence 6 by sandhi of "jīvan" and "eva", with duplication of the nasal "n" in the presence of a vowel, when it itself follows the vowel "a". We can clearly see here the comfort that brings us our segmentation software, which can recognize this thorny case.

Sentence 14 has a structure very characteristic of the story-telling style. it describes the succession of two actions performed by the same agent. The first action has a verbal form rendered by an absolutive, which is an invariable form of its verb. This absolutive clause precedes the main clause, which describes the second action with a conjugated verbal form. We follow the method from right to left as usual, from the red box tiṣṭhāma. By looking at its lemma, we see that it is an imperative of the first person plural of the root "sthā_1" (stay), here complemented by the instrumental sukhena from "sukha" (happiness) to express the meaning "let's be happy". It's about being happy as vibhaktāḥ , "those who have shared (the heritage)". The resemblance with "vibhāga", the sharing, which we saw above, is not by chance. "vibhakta" is the past participle of the verb "vibhaj" (to share), of which the noun "vibhāga" names the action. Going up the sentence, we find the absolutive clause, ended by the absolutive form gṛhītvā from the root "grah". The action is therefore to seize, and the agent is the "we" of the main clause. Seize what? The object awaits us, it is the compound vibhāga-dravyam in the accusative. The neuter substantive "dravya" is the substance of the first component, sharing. It is therefore the matter of inheritance. It is qualified by the initial compound tat-mañca-adhaḥ-sthitam formed on the past participle "sthita" from "sthā_1", so it is "what is under this bed". We therefore get literally "having taken ... let's be happy", which may be rendered as a conjunction: "Let us take the inheritance that lies under the bed, and after sharing it let us rejoice ".

The next sentence begins with a transition to story mode. The particle iti ends the speech of the previous sentence, which then serves as an argument to the absolutive uktvā which follows, to introduce the sentence with an absolutive clause: "having thus spoken". Then the following purple yāvat indicates a pair of co-relative clauses, the second starting at tāvat, co-relative of yāvat. Here co-relative means that the two pronouns co-denote, in this case expressing a logical implication. The sequence yāvat A tāvat B suggests that as soon as action A is completed, action B will occur. Here action A is for the brothers to dig under the bed, with a conjugated form of the root "khan" (to dig) which evokes cognate English "channel". Action B is a nominal judgment, whose predicate is nirgatāni, plural of the neuter past participle "nirgata" from the root "gam" preceded by the preverb "nis-", here in the sense of appearing. What things have appeared? tāmra-sampuṭāni: copper chests. How many? catvāri, four. Where did they appear from? adhaḥ, from below. Below what? caturṇām pādānām, the four-footed bed. With the genitive expected by adhaḥ, as already seen.

This four-footer may intrigue you, if you are not an Indian native. I translate it as "bed" in English, but in fact it is a platform where one can sit, that is to say sit cross-leg Indian style, have a meal, possibly lie down and take a nap. It consists of 4 sturdy wooden legs, carrying a frame stretched with wide strips of hemp fabric, forming a bed base. In the recent past, it was still commonly seen as furniture for dabhas, the traditional country inns for catering travelers. The Hindi word for them is charpad! Let us weep over Westernizing modernization, which is replacing the traditional charpads by ugly plastic chairs.

In the end, we rewind, and we unwind in contemporary English: "After these words, they dug under the bed to reveal four copper chests".

Let's go on. Sentence 16 begins with teṣām madhye: "from them in the middle". Here "them" are the chests which have just appeared and are the current topic. Then there is a locative clause ekasmin sampuṭe "in a chest". Here our attention is drawn to the three purple segments, which are variants of ekasmin: ekatra is "once" and anyatra is "another time". And in each of the four chests we are told what's inside, in the nominative: mṛttikā earth, aṅgārāḥ coals, asthīni bones, palālaḥ straw. We understand the construction, and we regurgitate: "One of these chests contains earth, another contains coal, another bones, yet another straw". We notice the "contains" which translates the asti implied main clause reduced to the content of each of the chests: mṛttikā is soil, but also means implicitly "soil exists". Our "contains" is necessary, because English requires a verb in each sentence, and even its subject, even fictitious, as in "it rains", while Sanskrit has no such restrictions.

Going to sentence 17, the red procuḥ jumps out at us. It is a form of the root "vac" (to speak) in the perfect tense, prefixed with the preposition "pra-" which is generally an intensifier, here meaning "they declared". "They" are of course the four brothers, te catvāraḥ , qualified over vismayam gatāḥ, literally "gone in astonishment". The two forms paraḥ param echo to say adverbially "each other". Everything is preceded by the absolutive clause governed by dṛṣṭva, "having seen". Having seen what? etat catuṣṭayam, this four-tuple (of chests). We recognize in catuṣṭayam a derivative of "catur", four, with a "-taya" suffix which can follow any number to denote the corresponding -tuple: "dvitaya" is a pair, "tritaya" a triad, and so on. We also notice that these two words are linked by sandhi with the "-t" end of "etat" assimilated to the initial "c" of the following word. In the end, we translate: "At the sight of the contents of these four chests, the four (brothers), flabbergasted, said to each other:".

Sentence 18 is therefore their exclamation, which begins with the interjection (green as a vocative) aho similar to our "oh". The sentence is here again predicative, governed by the past participle kṛtaḥ (done), qualifying vibhāgaḥ the share. The mauve pada samyak is an adverb meaning correctly, properly. The pitrā which precedes it is the instrumental of "pitṛ", which thus takes the role of the agent of this passive sentence. Indeed, while the subject (in the nominative) plays the role of the agent of a verb in the active voice, this is not so in the passive voice, where the agent is expressed in the instrumental, the nominative designating now the goal, here the sharing. The pronoun asmat is the ablative form of the personal pronoun "we", to mean "our". We recap: "Oh! Our father did the sharing fairly."

Sentence 19 is an interrogative in the passive, built on the root "jñā_1" (knowledge), whose agent in the instrumental is the interrogative pronoun kim: "How to understand this sharing?"

You have reached the middle of our story, at its climax. You built confidence, these passive constructions are now familiar to you, and you can rely on the computer tools to progress. Above all, you have not been put off by the acquisition of vocabulary, the richness of the declensions, or the difficulty of reading devanāgarī in continuous recitation, the segmenter-cum-tagger chewed up the work for you. Our method has saved you from the unrewarding drudgery of having to learn by heart the litanies of declension or conjugation who put off Latin students so much, and who for Sanskrit would be colossal, with 8 cases, 3 genders, 3 numbers, and countless verbal forms. Judge of it on the root grah : its forms run by the tens of thousands, each of which may also be prefixed by no less than 12 combinations of preverbs, leading to bizarre forms such as upasaṃjagṛbhuṣībhyām that you will never meet in texts, while very soon you will recognize the present gṛhṇāti he grabs, gṛhṇanti they grab, gṛhītaḥ seized, as well as to the normal degree form grahaḥ which is the grasper. Thus naming for instance the crocodile, but also the ladle, and sensory faculties such as smell or sight, in their role of sensors. In astrology, it is the role of celestial influence devolved to the Sun, the Moon, the five planets, the eclipse demon and the comet. These nine cosmic influences, navagrahāḥ (compound formed on "nava", the number nine), are carved on the lintel of the entrance to the sanctuaries.

By the traditional method, based on colossal memorization, it would have taken you months of hard work to read these few lines of text, while in less than an hour you have managed to grasp its meaning and on the way to familiarize yourself with the structure of the Sanskrit sentence. If you persevere in your effort, not only will you soon understand the enigma of sharing this bizarre inheritance, but you will easily acquire sufficient vocabulary to get by with simple texts. Courage!

We continue our reading with the 20th sentence. We find the iti uktvā already encountered, "having thus spoken". Notice that in the original text the sandhi welded the two vowels "i" and "u" to "yu". More generally, the final "i", "u" or "ṛ" changes on contact of a different initial vowel, and becomes the corresponding semi-vowel, respectively "y", "v" or "r".

This sentence has the now familiar structure of an absolutive clause giving the context together with a nominal main clause. But the absolutive form here does not end in "-tvā" like uktvā for the root "vac", but in "-ya" for the verb "upaviś", because this verb is prefixed with a preverb, here "upa-" meaning "near" or "under". We check in clicking on the purple box "upaviśya", we see that the verb is "upa-viś_1". By clicking on "viś_1" we find the dictionary entry of this root, its meaning (to enter, to settle), and the indication of its conjugation group [6]. By clicking on this 6 you visualize the page of its paradigms, where you will find in the invariable forms section the absolutive viṣṭvā, usable for this root stand-alone, but also "-viśya", which should be used when it is preceded by a preverb, such as "upaviś", hence upaviśya.

Coming back to the entry "viś_1" (assuming using the Heritage dictionary), you check that "upa" is in the list of preverbs that can be used with this root: (adhi, ā, upa, ni, pra, sam) and by clicking on "upa" you find the meaning of "upaviś", to sit. So upaviśya is "having sat". Where ? sabhām, meaning the hall of the city council, where the assembly is held. This accusative form has a feminine "sabhā" basis, as indicated by the long vowel "ā" which ends it. The genitive pronominal form tasyāḥ which follows refers to it; it is followed by the adverb puratas (in front), which precisely governs the genitive. The rest of the sentence consists of three padas in the singular nominative. Here the predicate is the past participle niveditaḥ of the causative of "ni-vid_1", verb meaning "to announce". And its subject follows it, this time. This illustrates the flexibility of Sanskrit, where the word order is not completely rigid. The pronoun form ayam is the demonstrative "this", qualifying the noun vṛttāntaḥ, meaning event. We therefore understand: "After these words they went to the council, whom they informed of this event".

The following sentence is simple. The predicate is jñātaḥ, participle of the root "jñā_1" (to know). You should pronounce "jñā" like "gnyā", thinking of cognate "gnosis". The na particle of negation indicates that a thing is not known. Which thing ? vibhāga-kramaḥ, the distribution of the inheritance. Who is the agent? The sentence being passive, the agent is in the instrumental, it is sabhyaiḥ, a plural form of "sabhya". This word is a secondary formation on "sabhā" with a suffix "-ya" meaning "concerning". We take it all back in the active voice, to keep our four brothers as the real agents, subjects of the verb to inform, because in the end if the council is in ignorance, it is that they did not explain: "They did not inform the members of the council of the distribution of the inheritance."

We then have a more complex sentence, with three verbs, that we shall have to dissect. It begins with the speech connective punaḥ, "however". Then after te catvāraḥ bhrātaraḥ (the four brothers) there is an incise: yatra yatra nagāre jñātāraḥ santi (anywhere in town there where knowledgeable ones), after which is the main proposition teṣām purataḥ nivedayanti (they informed), where of course "they" refers to the four brothers. The sentence ends with their statement, which begins by the pronoun "adas", , in its masculine accusative form amum. It is the distant demonstrative "that", contrasting with ayam already met, which is the close demonstrative "this" (masculine ayam, neutral idam, feminine iyam). The forms of these two pronouns, at the different genders, are varied but very frequent, and should be learned.