In an earlier article entitled Father sky in ancient Greek and Sanskrit the word designating the ancient Indo-European deity presiding over of the heavens has been discussed in connection with the very close correspondences between the declension of the word sky in ancient Greek and Sanskrit. The present article is its parent and discusses the declension of the word father in Greek and Sanskrit and the features which point to both words being sons of the same father.

Most people will know that all Indo-European languages have words for father that exhibit clear external features marking their provenance from one ancestral word. And it is indeed from the observation of the close relation of basic Sanskrit, ancient Greek and Latin cognates, such as पिता (pitā), πατηρ (patēr) and pater, that the entire discipline of comparative linguistics took its beginning. Although the relationship of the three forms above is obviously very close, it is also equally superficial – just an artificially immobile still of one moment in the development of a language or of one fixed shape that has been reached. But results in themselves are never of importance, the causes that led to the results are what matters – results are as mute as photographs, causes eloquent with tales of evolution. For more on the subject read The erosion of cases.

People are generally much less aware of the fact, that the most revealing insights about the relationship of the different forms of this very ancient word only comes once one explores their verticality, i.e. once one penetrates the crust of their outer appearance and compares the different case forms that constitute their filling. We do certainly not content ourselves with merely looking at a praline – no! we want to know what is hidden inside, be the surprise good or bad. So why be happy to look at the crust of words and not crack them open to reveal their declensional filling?

The forms discussed here are the nominative, accusative, genitive and dative of the singular. These are, in the order give above: पिता (pitā), πατηρ (patér);  पितरम् (pitaram), πατερα (patéra); पितुः (pituḥ), πατρος (patrós) and पित्रे (pitre), πατρι (patrí).

The first fact that becomes evident when one compares the different case forms of the words  पिता (pitā) and πατηρ (patér) is that the word consists of three components, two of which are modular, meaning they alternatively appear in longer and shorter forms. The characteristic feature of the nominative, both in Greek and Sanskrit, is the lack of the last of the three components, but they all appear clearly in the accusative, where we have पितरम् (pitaram) in Sanskrit and πατερα (patéra) in Greek, which words consist of the root पि-/πα- (pi/pa), the suffix -तर्-/-τερ- (tar/tér) and the case ending -अम्/-α (am/a). When one moves on to the genitive case of Greek, the same three components are still present, but here the repartition of weight between them has changed. In the genitive πατρος (patrós), the second element is suddenly contracted to a mere -τρ- (tr), while the case ending has the form -ος (os), the short form of which would be just -ς (-s).

Greek reveals that the factor governing this change between short and long forms of the last two components of the word is accent. In the accusative the accent is placed on the suffix -τερ- (tér) (note that accent is not indicated in the original Greek due to problems with the font used in this blog), while in the genitive the accent is placed on the case ending -ος (ós), what causes the suffix to assume its shortest form -τρ- (tr). The surprising fact is that the effects of these shifts of accent were so ingrained in the structure of the ancestral language from which Greek and Sanskrit later developed, that even in a language like Sanskrit, which has, in its classical form, lost all evidence of the original Indo-European accent system, these effects – the results of a lost cause – still resonate with great clarity in the shift between weak and strong cases, which is so distinctive of the classical language of India. So while the genitive पितुः (pituḥ) of Sanskrit is insufficiently explained and does not bear any clear relation to the Greek form given above, the dative पित्रे (pitre) has an exactly corresponding structure consisting of the root पि- (pi), the short form of the suffix त्र् (tr) and the long form of the case ending -ए (e). The Greek dative itself has the structurally exactly corresponding form πατρι (patrí), which consists of the root πα- (pa), the weak, unaccented suffix -τρ- (tr) and the strong, accented ending -ι (í). The reason for the absence of correspondence between the case endings of the Greek and Sanskrit forms, is that the Greek dative is often, as here, derived from the original Indo-European locative, which appears, in the case of the noun here discussed, as पत्रि (patri) or पितरि (pitari) in Sanskrit. The first form of the Sanskrit locative corresponds in every detail to the dative of Greek, the second form is a variant with accent on the suffix found in Sanskrit only.

What of the accusative? As we have seen above, it has the endings अम्/-α (am/a) in Sanskrit and in Greek. According to the rule that accented components appear in their long form and unaccented in their short form, the case ending should, due to the accent on the suffix, appear in its short form. But does it? What is the exact relationship of the two very different case endings of Greek and Sanskrit? The typical Indo-European marker for the accusative case is *-m (the asterisk indicates a form that has been reconstructed using the comparison of different attested languages, but is itself not attested as such). In stems ending in a vowel this resulted in the case ending *-om, for example, which is -ον (on) in Greek and अम् (am) in Sanskrit. When, as here, the stem ended in a consonant, the direct contact of the accusative marker resulted in the ending *-ṃ. The point under the letter indicates that the sound m functions as a vowel; that the m has become capable of bearing a syllable and can therefore act as a support for consonants, which can by themselves not constitute a syllable. Comparative linguistics has proved that the regular development of this syllabic m is -α (a) in Greek and अम् (am) in Sanskrit. The appearance of this set of sound correspondences must therefore mean that the ending in the accusative was originally *-ṃ, meaning that it assumed, as expected, its shortest form due to the absence of accent on it.

It now becomes clear how little information is to be gathered from the nominative alone and how little it tells us about the depth of the relationship between the two words as a record of their growth and development; how little it tells us about words seen not as inanimate shapes on paper, but as living entities that grow, change and breathe the same air than we do. But what of the nominative? In both languages the nominative is characterised by the absence of a case ending, which is replaced by the special lengthened grade of the suffix. According to this पिता (pitā) consists of the root पि- (pi) and the lengthened suffix -ता (tā), the Greek word πατηρ (patēr) of the root πα- (pa) and the lengthened suffix -τηρ (tēr). It is a typical feature of Sanskrit that this kind of suffix loses its final consonant when appearing in its lengthened form. For example Sanskrit माता (mātā) corresponds to Greek μητηρ (mētēr) and the nouns राजा (rājā) or आत्मा (ātmā) are n-stems, which have the base forms राजन् (rājan) and आत्मन् (ātman).


Silvio Zinsstag,

teacher for ancient languages