Version of 2016-09-23

Wersja polskaBilanguage versionWersja dwujźzyczna

Previous part Contents Next part

Grzegorz Jagodziński

Indo-European and Semitic languages

1 – 2 – 3

Expressing of possessing

Older languages of both analysed families have a special case – genitive. It should be emphasized that some languages of other families, also these, in which the category of case exists, often express possessing in a different way, for instance with the construction this man wife-his ‘the wife of this man’, in which both the qualified and qualifying words appear in the same case. Certain differences can be seen in the order of both elements. In the Semitic languages the qualified word stands, as a rule, before the qualifying word (which appears in genitive), e.g. house large, house brother’s ‘a large house, brother’s house’. In the Indo-European languages the reverse sequence prevails when a substantive is determined with an adjective, e.g. large house, but when a substantive is determined with another substantive, the sequence is often the same as in Semitic, e.g. house of (my) brother (cf. the synthetic form in Polish: dom brata).

Genitive in some IE languages can have a special suffix, originally a word-formation suffix, which is in Latin substantives which end in nominative with -us or -um (from older -os, -om), e.g. templum ‘temple’ – templī. The same suffix can be found in the Semitic languages – it forms, among others, adjectives from substantives as well as names of inhabitants from country names (cf. also the genitive ending -i).

Sometimes the IE genitive has the ending taken from another case, e.g. Polish -a in wilk – wilka ‘wolf – of the wolf’, which is originally an ending of ablative -ōd, the case of distance (where from? from whom? from what?). However, there exists a theory saying that the original method of building genitive was a change of the stress place from the root to the ending. Next, as the result of the subsequent reduction of unstressed syllables, a secondary genitive ending developed. Let’s consider the instance of the IE substantive with the meaning ‘foot’:

It is possible that a vowel alternation accompanied the presented changes. Some data suggest the possibility to reconstruct the forms of the A phase as *pódos – *pedés. When the substantive had a suffix with a vowel, the changes of that kind led to other types of apophony, because the stress shifted not on the ending but on the syllable which followed the root: N *ghost-is ‘visitor, alien, guest’ < *ghóstejes, G *ghosteis < *ghostéjes.

Originally only 3 cases existed in the Semitic languages. They were formed with the help of the following endings: -u for nominative, -i for genitive and -a for accusative. The so called mimation -m, which expressed indefiniteness, can have followed the case endings in the most nouns which are called nomina triptota. There existed also a special of nouns which are called nomina diptota. They could not have the mimation and they had -a in indefinite genitive (like in accusative). Such a system is preserved in the literary Arabic, in which the mimation has changed into nunation (-n). Cf. farasun ‘a horse’ and ˀāḫaru ‘other’:

  nomen triptotum nomen diptotum
nominative farasun, al-farasu ˀāḫaru, al-ˀāḫaru
genitive farasin, al-farasi ˀāḫara, al-ˀāḫari
accusative farasan, al-farasa ˀāḫara, al-ˀāḫara

At the same time they preserved the separate ending -i in definite genitive. If the substantive was qualified by neither an adjective nor a different substantive, it was in the form called status absolutus. When the substantive was qualified by an adjective or another substantive, it was losing the mimation, the stress, and it was in a special form called status constructus (Aramaic has also status emphaticus with the ending which is of unclear origin). E.g.:

It is worth to be noticed that the stress has moved onto the substantive in genitive. The forming of genitive in IE by changing the stress place onto a next syllable is not a strict equivalent of the Semitic construction, however it is possible to see some connection between them.

Gender

The category of gender is not too universal in languages in the world. It is quite inversely – we will fail to find genders, for instance, in Uralic or Altaic languages. Indo-European and Semitic (and also other AA languages), however, have such a category, which make them similar in this point.

Indo-European languages

It has been stated that in Proto-IE 3 genders existed: masculine, feminine and neuter. Many daughter languages, including modern languages, inherited that feature. The 3 mentioned genders appear in Sanskrit, Greek or Latin, and also e.g. in German.

Gender is a feature which is attributed to each substantive. In languages, in which the category of gender exists, adjectives (and words with similar function, like some pronouns and numerals) appear in various forms depending on which gender is the substantive which is qualified by them. Here is a table showing the phenomenon of gender (a copy from the page devoted to the Polish grammar):

  gender
masculine feminine neuter
Polish mój dobry brat moja dobra siostra moje dobre dziecko
Latin meus bonus frāter mea bona soror meum bonum (opus)
German mein guter Bruder meine gute Schwester mein gutes Kind
English my good brother my good sister my good child

In some IE languages the category of gender changed. For instance, Romance languages lost neuter gender – hence they have only masculine and feminine genders. Admittedly, English shows traces of old genders in the forms of the pronouns he, she, it which correspond to 3 old genders, however beside this distinction genders have disappeared here completely – and that is why adjectives or articles appear in only one form, independently on the word which they qualify.

In Polish the number of genders has increased thanks to development of the categories of animateness and personality. Additional differences are noticeable in nominative plural and in accusative of both numbers (brat – brother, pies – dog, wóz – cart).

  gender
masculine personal masculine animate masculine inanimate
N sg. mój dobry brat mój dobry pies mój dobry wóz
A sg. mojego dobrego brata mojego dobrego psa mój dobry wóz
N pl. moi dobrzy bracia moje dobre psy moje dobre wozy
A pl. moich dobrych braci moje dobre psy moje dobre wozy

In Polish or German nearly every adjective is able to build separate forms for particular genders. However in older IE languages it was not always so. For instance in Latin we must say meus mināx frāter, mea mināx soror, meum mināx opus (‘my menacing father’, ‘my menacing sister’, ‘my menacing work’) – how it can be seen, some adjectives in this language do not distinguish genders. Sometimes only two gender forms appear – one, named utrum, for masculine and feminine genders, and the other, neutrum, for neuter gender. For instance meus juvenis frāter, mea juvenis soror, but meum juvene opus (juvenis – ‘young’). We can see a similar situation in Greek, which is also mentioned below.

As the oldest known IE languages we can number the Anatolian languages, and among them Hittite and Luwian. In those languages feminine gender is completely absent, and the difference between utrum (or genus commune – the common gender) and neutrum (the neuter gender) exists. A characteristic feature of the neuter gender is one form for both nominative and accusative, while in the common gender (similarly like in masculine and feminine in those languages where such appear) are usually present different forms of either cases. Another interesting phenomenon is also easy to be perceived. And so, nouns of the neuter gender of the -a- (< IE *-o-) declension have the ending -n < *-m in nominative–accusative, the same as in accusative of the common gender of this declension. At the same time neuter nouns of other types of declension have no ending in nominative–accusative unlike accusative of the common gender. E.g. in Hittite:

An identical phenomenon can be observed in Sanskrit, Greek, Latin etc, when comparing the declension in masculine and neuter genders, e.g. Lat. the neuter substantive templum ‘temple’ has the ending -um in NA, the same as in accusative of a masculine substantive (e.g. amīcum from amīcus ‘friend’), but the neuter substantive mare is the naked root without an ending, in contrary to accusative of a masculine substantive of the same type (e.g. piscem from piscis ‘fish’). A comparison with the division of Semitic nouns into nomina triptota with the mimation and nomina diptota without it imposes itself.

Hittite common N atta¹, A attan – neuter NA pedan, common N ¹alli¹, A ¹allin – neuter NA ¹alli,
Latin m N amīcus, A amīcum – neuter NA templum m N piscis, A piscem – neuter NA mare
Arabic triptotum farasun, al-farasu diptotum ˀāḫaru, al-ˀāḫaru

In IE languages there are not universal markers – suffixes which mark gender. It can be observed that feminine substantives usually are marked some special suffixes while masculine and neuter substantives need not have any special endings. In the Polish language adjectives have strictly determined endings for individual genders (-y or -i for masculine, -a for feminine, -e for neuter), but a similar rule cannot be made for substantives. Many masculine substantives end with a consonant, many feminine ones end with -a, and many neuter ones end with -o, -e or , but it is not a firm rule. For instance

There is even worse in Latin. Here none strict rules cannot be found also for adjectives. Masculine gender is often marked with the ending -us, feminine with -a, neuter with -um, but there also exists a fair group of adjectives with different gender endings (see the examples above).

In Greek, adjectives with three endings end with -os, -ē (-ā), -on (they strictly correspond with Latin -us, -a, -um), however there exist adjectives which end with -os in masculine which have not the feminine form (and which are using the masculine form for feminine). Compare:

In modern languages the largest differences exist between the feminine gender and the others – for instance masculine and neuter genders have identical forms in most of the oblique cases in Polish. Such a situation can cause neuter gender to disappear in some languages. On the other hand, in older languages differences between neuter and the two other genders were more important. And so, in neuter there did not exist separate forms of nominative and vocative – both had the same form as accusative. Adjectives with two endings always had one separate form for the neuter gender, and another for both masculine and feminine. The nominative ending -s (called the sigmatic ending) never marked neuter gender – cf. Latin juvenis, juvenis, juvene ‘young’ or Greek glykżs, glykeĩā, glykż ‘sweet’. Forms of the type of Latin mināx were transferred secondary into neuter gender, and they replaced older *mināc because such a form was rare due to its meaning (similarly on Greek pénēs ‘poor’ which is an adjective with one ending and similar instances of adjectives which were practically unused in neuter).

Sanskrit lets us make other interesting observations. The Latin endings of many adjectives -us, -a, -um and Greek -os, -ē (-ā), -on have their counterparts as -as, -ā, -am in Sanskrit (because of sandhi the final -s often appears as -ḥ and hence many sources gives -aḥ, -ā, -am), however the feminine form ends with , not with , in a great number of instances. For example

Substantives which denote persons behave similarly: devas means ‘god’, devī – ‘goddess’, also kumāras – ‘boy’, kumārī – ‘girl’. It can even be stated that the ending plays the same general feminizing role in Sanskrit like -a in the Polish language – that is why the feminine form balinī was made as the counterpart to masculine balī < *balins ‘strong’ and neuter bali, balin. The vowel -i- enters into some suffixes even when the ending is , e.g. pācakas ‘he-cook’, pācikā ‘she-cook’.

Traces of the feminine suffix -ī, -īk, -ikā are also preserved in other languages: hence Polish bóg – bogini ‘god – goddess’, pan – pani ‘mister – mistress’, but also Latin inventor – inventrīx (the same their Polish counterparts wynalazca – wynalazczyni). The suffix seems to be more archaic than which spread in newer formations. It could also be related to other feminine endings: Lithuanian and Latin -ē-s.

It is very characteristic that nominatives of the forms with the suffixes -ā, -ī was asigmatic, i.e. it did not come together with the regular endings -s (however cf. several Sanskrit exceptions like lakṣmīs ‘hapiness, good luck’, tantrīs ‘string’ as well as Latin -ēs, btw. with uncertain origin). It should be supposed that their original form were -aə, -iə < -aH, -iH. The second suffix, , was built of two sonants, hence it was able to appear as -jə in some instances (originally after one consonant). We can see this form in the Greek words moũsa, moĩra < *montjə, *morjə. In Sanskrit there also exists a group of feminine words which are affixed with . Their nominative is sigmatic, e.g. vadhūs ‘bride, fiancée’, ¶va¶rūs ‘mother-in-law, mother of the husband’.

Finally, it should be emphasized that feminine suffixes seem to build neuter nominative plural forms. Originally it was expressed with a collective substantive and that is why this form was treated as singular in syntax. The most common marker of neuter plural is , which can be seen e.g. in Slavic languages (Pol. -a in s³owa, imiona, cielźta ‘words, names, calves’). However, in Greek and Latin we can find the short -a, and in Sanskrit we can find -i, which derives from IE *-ə < *-H. Cf. also Vedic vanā (Skr. vanāni, sg. vanam ‘water’, an -a- stem), vārī (Skr. vārīṇi, sg. vāri ‘water’), madhū (Skr. madhūni, sg. madhu ‘mead’), with long vowels from groups with a laryngeal.

The Semitic languages

A characteristic feature of all Semitic languages is the presence of 2 genders: masculine and feminine. Only few feminine substantives lack morphological markers (feminine suffixes). In Arabic names of body parts belong here, e.g. ˁajnun ‘eye’, ˀuḏnun ‘ear’, jadun ‘hand’, riglun ‘leg’, names of some animals, e.g. ḍabuˁun ‘hyena’, ˀarnabun ‘hare’ and few other names, e.g. ¹amsun ‘sun’, nārun ‘fire’.

There exist several suffixes which are used to create feminine forms.

  1. The suffix -t, e.g. Arab. ˀuḫtun ‘sister’ (ˀaḫun ‘brother’), bintun ‘daughter’ (ibnun ‘son’).
  2. The secondary suffixes -at, -it, -ut, which were made by joining -t and word-formation elements.
  3. It is controversial whether the suffix or -ah (for feminine gender) ever existed in Semitic. However, it should be noticed that in Arabic the ending of most feminine substantives -atun is not written with t, but with the specially modified letter h. Instead of -atun in writing, the pronunciation is present today, and this phenomenon seems to be very old. There is similarly in Hebrew, where from the Biblical times we have -āh in spelling and in pronunciation instead of the expected *-āṯ. It is interesting that in Aramaic two different groups of feminine substantives appear. In the first group there are st. abs. , st. constr. -aṯ, st. emphat. -tā, while in the second group there are st. abs. = st. constr. -āṯ and st. emph. -āṯā, e.g. məḏīnā, məḏīnaṯ, məḏittā ‘country’, but mənāṯ, mənāṯ, mənāṯā ‘lot, share’. Maybe we can find the old suffix -ah alternating with -t in the 1st group, while the composite suffix -at in the 2nd group.
  4. The suffix -āˀ, present e.g. in some Arabic adjectives, like ḥamrāˀu ‘red (f)’ (aḥmaru ‘red (m)’). Such forms are nomina diptota, hence they lack the nunation (-n). The same suffix -āˀ is used to create some plural forms, e.g. ¹uˁarāˀu ‘poets’ (sg. ¹aˁirun), aṣdiqāˀu ‘friends’ (sg. ṣadīqun). In Ethiopian (Geez) that suffix occurs (as the ending ) in some abstract names (ḥanṣā ‘building industry’) and concrete ones (ḥamadā ‘snow’). in Hebrew its counterpart is the ending that occurs in town names (Hebrew -ō- is the result of a regular phonetic change of the Semitic -ā-; this vowel is also spelt -ō- since it might have been pronounced in another way than the ordinary -ō- < -u-, as well as because of the peculiarities of Hebrew orthography).
  5. The suffix -āj occurs in Arabic ( in pronunciation, -aj in spelling). This suffix forms feminine forms of elative, e.g. ḥusnā(j) ‘very beautiful (f)’ (masculine aḥsanu ‘very beautiful’, the positive grade ḥasanun ‘beautiful’). The adjectives which have the suffix -ānu (nomina diptota, the same as the elative forms) also build their feminine forms with -āj: kaslānu ‘lazy (m)’, kaslā(j) ‘lazy (f)’. Finally, feminine substantives with -āj form action names with passive meaning: ḏikrāj ‘recalling’. In Ethiopian the analogical suffix occurs in such substantives as ˀarwē ‘beast’, nāgē ‘elephant’, ¹arwē ‘tree trunk’, sarwē ‘army’, gīzē ‘time’, massalē ‘comparison’. In Hebrew and Aramaic suffixes -aj, -ē appear, but they can also form masculine names, e.g. hbr. ˀarjē ‘lion’, ḥōraj ‘fabric gender’, Aram. ˀoḥrē ‘other (f)’.
  6. The suffix (-i) seldom occurs singly: in Arab. infinitives which are used as interjections, e.g. nazāli ‘come down!’, tarāki ‘leave it alone!’, in Arab. interjections which are used towards women, e.g. ja fagāri ‘oh unruly woman!’, lakāˁi ‘oh wicked woman!’. In other instances the original was developed to the form -ijjatun. In Aramaic, occurs only in status absolutus forms of some substantives, e.g. debbōrī ‘bee’; in other forms the common feminine suffix -t is added. No one knows whether we can observe an instance of reduction of the original -t, or the primitive form is preserved here.
  7. The suffix occurs in Aramaic in abstract names in status absolutus, e.g. tajbū ‘right’ (st. constr. tajbūṯ, st. emph. tajbūṯā), malkū ‘State, politic body’. Furthermore, the Semitic languages know only -ūt, which can show secondary development. Examples: Akk. bēlūtu ‘State’, hbr. malkūṯ ‘t.s.’, Arab. malakūtun.

Some Hebrew facts are especially striking. Namely, many Hebrew masculine substantives with the zero ending have feminine counterparts with the ending , which delusively resembles IE facts, like those present in Polish: pār ‘bull’ – pārā ‘cow’; ˁēḡel ‘he-calf’ – ˁeḡlā ‘she-calf, young cow’, sūs ‘horse’ – sūsā ‘mare’.

Previous part Contents Next part