dhAtupATha

The 'dhAtupATha' ( धातुपाठ ) is a text that is closely associated with the Ashtadhyayi 1 The following notes are intended to help the reader decode the entries in the 'dhAtupATha' (clicking on this link will take you to an external site that contains the complete text of the 'dhAtupATha' [1]).

The 'dhAtupATha' contains a list of all the verb classes (1-10) and the verb roots in each class along with a semantic mnemonic that indicates its semantic equivalence class. For e.g., the first few entries are the following:

अथ॒ भ्वा॑दयः ।
भू सत्ता॑याम् । उ॒दात्तः॒ पर॑स्मैभाषः ॥
अथ॑ तव॒र्गीया॑न्ताः ।
एधँ॒ वृद्धौ॑
...
इत्येधा॑दय उ॒दात्ता॒ अनु॑दात्तेतः ॥

• The first line above tells us that the 'bhvAdaya' class starts here (i.e. verb class 1 'bhU, etc.'). This verb class continues until the onset of the next class is signalled by a similar statement. The statement अथादा॑दयः (i.e. verb class 2 'ad, etc.') can be seen after over a thousand entries (verb class 1 accounts for a little over half the entries in the 'dhAtupATha').
• The highlighted segments in the above entries tell us that the verb root 'bhU' (in verb class 1) belongs to the semantic equivalence class 'sattA=existence/being', while the verb root 'edh' belongs to the semantic equivalence class 'vRuddhi=increase'.
• Please note that several verb roots may share the same semantic mnemonic. For e.g., 'vid' in verb class 4 shares the semantic mnemonic 'sattAyAm' with 'bhU'. Similarly, 'pyAy', 'maNh', 'pUSH', 'bRuNh', 'dakSH', 'pyE' (all in verb class 1), 'Rudh' (verb class 4), 'hi', and 'Rudh' (verb class 5) share the semantic mnemonic 'vRuddhO' with 'edh'.
• Note also that some verb roots may specify multiple semantic mnemonics on the same entry (for e.g., 'dakSH' in verb class 1 adds the mnemonic 'shIghrArthe=quickness' with the coordination marker 'cha'). Sometimes, there may be an additional semantic mnemonic followed by 'ityeke' (which can be read as 'alternately ... as per others').
• Some entries may not specify the semantic mnemonic and end in the 'avagraha' symbol ( ऽ ). These usually share the semantic mnemonic of the next entry that specifies the semantic mnemonic. Hence, for e.g., the entries कुर्दँ॒ऽ and खुर्दँ॒ऽ in verb class 1 tell us that 'kurd:to play' and 'khurd:to play' share the semantic mnemonic 'guDakrIDAyAmeva:playing ball' with the next entry गुर्दँ॒ ('gurd:to play').
• Note also that each verb root may have multiple semantic equivalence classes. For e.g., the verb root 'vid' has 5 different semantic classes (with different accented verb roots), corresponding to the verb classes 2, 4, 6, 7, and 10. These must be treated as independent verb roots (polysemy) as they have different syntactic as well as semantic properties. Hence, a syntactic analysis will mention the verb root as 'vid:2:to know', 'vid:4:to be', 'vid:6:to find', etc..
• The verb root is listed using an accented form (for e.g., एधँ॒ and अ॒दँ ) with 'it'-markers that indicates certain syntactic properties used in its conjugation (for e.g., 'it'-markers (such as the word-final अँ in एधँ॒ and अ॒दँ ) and Accent-markers (such as the horizontal line below some letters in एधँ॒ and अ॒दँ ). The significance of the Accent-marker and 'it'-marker is explained in the Notes that follow.
• Some key syntactic properties mentioned in the 'dhAtupATha' include:
• the word-final 'it'-marker
• other 'it'-markers
• the Accent-marker for the main vowel
• the Accent-marker for the word-final 'it'-marker
• the Voice (Parasmaipada, Atmanepada, Ubhayapada)
• The verb class (1-10)
• the verb grouping that sometimes attracts special sutras during verb conjugation or derivative formation. For e.g. the 'rudAdaya' grouping in verb class 2. These groupings may mention the Voice and Accent applicable to all members of the group, in which case this will override the Voice or Accent derived from any individual accented entry (if it is different). Hence it is advisable to scroll down to the end of each verb grouping to see if any overarching Voice/ Accent is specified for the entire group.
• The entry on the first line that says that 'bhU' has the Parasmaipada Voice and the 'udAtta' Accent-marker (this is the accent of the main vowel as well as the 'it'-marker). This classification is usually specified at the beginning of each grouping, and applies to all the entries following until the end of the grouping. The grouping usually ends with a statement that says 'iti ...' or a statement indicating a different Voice/ Accent, or the beginning of a new grouping.
• In the case of 'edh', it is in a new grouping preceded by the statement that says the 't'-ending verb roots will follow ( अथ॑ तव॒र्गीया॑न्ताः ). This grouping is termed as 'edhAdaya' at the end of the grouping (in the final statement of the grouping i.e. इत्येधा॑दय उ॒दात्ता॒ अनु॑दात्तेतः ), and specifies that all entries of the group have the main vowel Accent marker of 'udAtta' (i.e. they take the 'iT'-augment), while they all have the 'it'-marker 'anudAtta' ('anudAttetaHa' i.e. they all have the Atmanepada Voice).

Notes on it-markers:

• The verb root must be extracted after dropping all special markers (i.e. 'it'-markers as well as Accent markers). However, note that 'it'-markers and Accent markers are important syntactic attributes, and may attract special sutras in some conjugations.
• It must be noted that multiple 'it'-markers may be present, both before and after the verb root, and the extraction of the 'dictionary' form may occasionally require some analysis. Some of the prefixes include:
• Word-initial ष will result in extracted verb roots with word-initial स instead. Hence, from षु॒ञ् we obtain सु.
• Word-initial डु is treated as an 'it'-marker that must be dropped. Hence, from डुक्री॒ञ् we obtain क्री.
• Word-initial टु is treated as an 'it'-marker that must be dropped. Hence, from टुक्षु we obtain क्षु.
• Word-initial ञि is treated as an 'it'-marker that must be dropped. Hence, from ञिभी॒ we obtain भी.
• Not all verb roots have a word-final 'it'-marker, as can be seen in the case of 'bhU' in verb class 1.
• अँ (52%): This ending is the common case and requires no special treatment (other than the dropping of all markers).
• इँ (13%): This ending indicates a 'num' dhAtu (i.e. it attracts the {7.1.58: idito num dhAtoHa} sutra). This sutra says that the nasal 'n' must be inserted after the main vowel. Hence the verb root 'skund' will be extracted from the accented verb root स्कुदिँ॒ .
• इँर् (1%): This ending usually indicates the 'iT'-augment in some conjugations. For e.g., रुदिँर् , listed in the 'rudAdaya' grouping in verb class 2 attracts the {7.2.76:rudAdibhyaHa sArvadhAtuke} sutra, where conjugational affixes that begin with a consonant take the 'iT'-augment, resulting in, for e.g., 'roditi' for the verb root 'rud:2:to weep'. To extract the verb root, the markers may simply be dropped (hence, we obtain the verb root 'rud' from रुदिँर् ).
• ॠँ (8%): This ending indicates the applicability of a special conjugation sutra. To extract the verb root, the markers may simply be dropped (hence, we obtain 'bAdh' from बाधृँ॒ ).
• उँ (5%): This ending indicates the applicability of a special conjugation sutra (application of the 'iT'-augment). To extract the verb root, the markers may simply be dropped (hence, we obtain 'aNch' from अञ्चुँ ).
• ऊँ (1%): This ending indicates an optional 'iT'-augment in some conjugations. Hence, the infinitive for the verb root 'gup:1:rakSHaNNe' ( extracted from गुपूँ ) may be either 'goptum' or 'gopitum'.
• ङ् (2%): This ending indicates the Atmanepada Voice. See, for e.g., sU ( षूङ् ) in class 4 in the table.
• ञ् (2%): This ending indicates the Parasmaipada Voice. See, for e.g., vRi ( वॄञ् ) in class 9 in the table.
• There are many more such special 'it'-markers. However, the above markers cover around 84% of all verb roots. Verb roots that have a vowel-end and no word-final 'it'-marker (such as the first entry 'bhU') account for an additional 7% of verb roots. But, in general, any other word-final 'it'-marker that consists of a vowel+'anusvAra' (for e.g., ईँ , एँ , ओँ , लृँं ) can be dropped in order to extract the verb root. These 'it'-markers serve to attract special sutras during some conjugations or formation of derivatives.

Notes on Accent Markers:

• A horizontal line below a letter (for e.g., अ॒दँ ) indicates the 'anudAtta' accent (Greek 'grave'), while a vertical diacritic on the right of a letter (for e.g., द्विषँ ॑ ) indicates the 'svarita' accent (Greek 'circumflex'). The accent is 'udAtta' (Greek 'acute') in the absence of either of these signs (unless overridden by a verb grouping statement, as in the case of 'hu' shown in the table below).
• The Accent Marker for the word-final 'it'-marker tells us the Voice of the verb root. The 'anudAtta' marker indicates Atmanepada, the 'svarita' marker indicates Ubhayapada, and the 'udAtta' marker indicates Parasmaipada. We can see examples of these in the class 2 verbs ('ad': अ॒दँ , 'dviSH': द्विषँ ॑ , and 'ID': ईडँ॒ ) shown in the table.
• The accent of the main vowel tells us whether the verb takes the 'iT'-augment in some conjugations. For e.g., the affix is 'itum' for infinitives for verbs where the 'iT'-augment is required, and 'tum' where it is not required. In a few cases, the 'iT'-augment is optional, but these cannot be determined by the accent alone.
• seT (sa+iT: with 'iT'-augment): If the main vowel has the 'udAtta' accent, as seen in 'edh' ( एधँ॒ ), then the 'iT'-augment is required (resulting in the infinitive form 'edhitum').
• aniT (an+iT: without 'iT'-augment): If the main vowel has the 'anudAtta' accent, as seen in 'ad' ( अ॒दँ ), then it does not take the 'iT'-augment (resulting in the infinitive form 'attum').
• veT (vA+iT: optional 'iT'-augment): These are not indicated by the accent (they will instead have the 'udAtta' accent, indicating 'seT'), but instead are specified by various sutras in the Ashtadhyayi. For e.g., 'sU' (shown in the table below), is mentioned in the {7.2.44: svaratisUtisUyatidhUJUdito vA} sutra, along with 'dhU:kampane', 'svRu:shabdopatApayoHa', and all accented verb roots ending with the 'it'-marker ऊँ (see the entry for ऊँ in the preceding Notes on 'it'-markers ).

The verb roots shown in Roman letters in the table below are in the 'dictionary' form, without 'it'-markers and Accent-markers (for e.g., 'edh' and 'ad' are extracted from एधँ॒ and अ॒दँ ). The accented verb root, from which the verb root is extracted, is shown in brackets (please look closely at 'ash' and 'man' where a diacritic is overlaid on top of the horizontal accent marker). For each verb root, the following table lists the commonly associated Voice ('P:parasmEpada', 'A:Atmanepada', 'U:ubhayapada'), Accent ('ud:udAtta', 'an:anudAtta', 'sv:svarita'), 'iT'-Augment ('S:seT', 'A:aniT', 'V:veT'), and a sample conjugation:

A few examples from Class 1 ('bhvAdi') are listed below:

A few examples from Class 2 ('adAdi') are listed below:

A few examples from Class 3 ('juhotyAdi') are listed below:

A few examples from Class 4 ('divAdi') are listed below:

A few examples from Class 5 ('svAdi') are listed below:

A few examples from Class 6 ('tudAdi') are listed below:

A few examples from Class 7 ('rudhAdi') are listed below:

A few examples from Class 8 ('tanAdi') are listed below:

A few examples from Class 9 ('kryAdi') are listed below:

A few examples from Class 10 ('churAdi') are listed below:

• 1. It must be noted here that there are several versions of the dhAtupATha (associated with various schools or commentaries), containing over 2,200 verb roots. However, lexicographers have observed that nearly half of these verb roots are not to be found in any text that has survived to the present day, an indication of the quantum of literature that has been lost over the last two millennia from Panini's era (c. 500 BCE).

References