Palmer L.R. The Greek Language

Подождите немного. Документ загружается.

< previous page page_259 next page >

Page 259

Compounds may also be classified according to the syntactical relation of their members. This criterion yields two

classes: 1, those in which the two members are in apposition, e.g.

'a day and a night' (this type is rare

in ancient Greek but is more common in modern Greek, e.g.

'a married couple').; 2, the determinative

compounds, in which normally the second member is determined by the first (the type

'river-horse'

being late in Greek). There are three main sub-types:

(i) Adjective plus noun:

'upper city'.

(ii) Noun plus noun: 'father's brother'. When the second member is an agent noun, the compound

corresponds to a verb plus a direct object:

'father-slayer'.

(iii) The first member is a verbal stem governing the noun stem that forms the second member:

'beginning evil', 'enduring work', 'fleeing war', 'loving war',

'carrying house' (i.e. a snail).

In some compounds of this type the first member is distinguished by the suffix which is used to form verbal

nouns (see p. 252): 'joining words', 'glib', 'feeding man', 'rousing the army'.

This suffix underwent assibilation in the East Greek dialects (p. 60), e.g.

. However, not all the first

members which have

are traceable to this origin. In fact the greater number appear with verbs which have

sigmatic features, desideratives or aorists, in their paradigms:

'persuading mortals' may serve as an

instance, for the verbal noun is < *bhid-ti-. Other examples are 'raising the feet',

'delighting mortals', etc.

To turn now to the form of the compounds, we distinguish between the first and second members. The first may be

(a) the pure nominal stem, e.g. . Ablauting stems tend to have the weak grade: e.g. 'falling on

the knee', 'winter flowing', where the -a- < * ; (b) a case form: e.g. 1 'Lord of the Earth'

(vocative, cf. Lat. Jupiter, *

), (genitive), (locative),

(instrumental); (c) one of a number of inherited elements, notably the negating

1 This is the presumed basis of the name Poteida(h)on, Poseidon.

< previous page page_259 next page >

< previous page page_260 next page >

Page 260

prefix

(e.g. ) and the so-called 'copulative' prefix, which in fact is the zero grade *sm- * of the word for

'one' *sem-. This would regularly appear as

( , etc.), but lost its breathing according to Grassmann's

Law (p. 230) when the second member contained an aspirate: , (sm-gwelbho-s*), , etc.

The unaspirated form spread by analogy to compounds where no aspirate followed. There was, of course, no

distinction in the dialects which lost the aspiration.

Other first elements are:

;

(d) prepositions: , etc.; (e) adverbs (often fossilized noun cases (p.

283)):

; (f) cardinal

numbers: ; 'having four , appears to have a zero form of the word for

'four' *qwtru-*, but cf.

and see p. 290.

A large proportion of first members were noun stems of the thematic declension whose stems ended in -o-. As a

consequence this vowel was regarded as a composition suffix par excellence and became attached to other stems:

(earlier ).

Turning now to the form of the second members, we find in this position a certain number of stems which do not

occur as separate words and also do not appear as first members. A number of these are ancient root nouns which

have survived only in composition:

, etc. More usually such stems are extended

by means of a compositional suffix:

(*o-sd-o-s, -sd- being the zero grade of sed- 'sit').

Another ancient compositional suffix is -t- (cf. Lat. sacer-do-t), which appears in , , etc. In IE -

a- also functions as a compositional suffix (cf. the Latin type agri-col-a) as in the ancient word , the

first member of which is traced to *dems-, perhaps a genitival form of the word for 'house'; -a- may be added to -t-

, -ta- being used to form compound agent nouns like

( < *-ed- 'eat' + ta; for

*dt > st, see p. 241).

< previous page page_260 next page >

< previous page page_261 next page >

Page 261

B. Verbs

The Greek verbal conjugation comprises a number of different 'tense' stems, present, aorist, perfect and future. The

first three were inherited from IE, but the future in all IE languages is a recent formation (see pp. 301f.). These

stems have distinctive Ablaut grades: e.g., present

, aorist , perfect . Although they are called

'tenses', the stems have no temporal references but relate to the type of action. The so-called present stem

represents continuing action (the durative stem), the aorist refers to momentary action, while the perfect represents

the state resulting from an event. Thus

means' to be in the process of dying', 'to expire' and

'to be dead'. These different presentations of verbal events, durative, aoristic, and perfect, by distinctive stems are

known as the 'aspects' of a verb. By different devices, which will be discussed below, the aspectual stems give rise

to a variety of 'moods' indicative, subjunctive, optative, imperative and infinitive. In the indicative, temporal

distinctions can be made by means of the augment and inflexions: thus from one and the same durative stem,

, we may form a present tense and a preterite .

The description of the function of the 'present' stem as 'durative' does not cover all the facts. Thus when Socrates'

friends visited him in prison they entered and 'found' him just relieved of his shackles. Nothing could be more

momentary than this act of 'discovery'. Yet Plato uses the durative form

. This has greater directness

and vividness: it represents the event as it unfolds as in a news-reel. We have called it the 'eye-witness aspect'. The

aorist, by contrast, simply refers to the event as a unit of history (p. 293).

At this point only the formation of the 'present' stem will be under consideration. Although Greek inherited from

IE a variety of present-stem formations, most of these survived as mere fossils. The productive types, which give

the Greek verbal system its peculiar stamp, in the main owe their origin to a single inherited suffix *-je/o-, which

could be added to both vowel and consonant stems. The exposition thus falls naturally into two sections (A) the

fossils and (B) the -j- formations.

< previous page page_261 next page >

< previous page page_262 next page >

Page 262

1. Root stems. These are divided into two groups according to the presence or absence of the 'thematic vowel', a

fundamental phenomenon in the Greek verbal conjugation, which may be illustrated from the contrast of

. In the first member of these pairs a stem-forming ('thematic') vowel e/a appears

before the inflexion, but not in the second member. We have, then, (a) thematic root stems:

, (*segh-

), , etc.; and (b) athematic: (*esmi), , (Mycenaean, later replaced by ), , etc.

2. With reduplication and zero grade: (a) thematic: (cf. (Lat. gi-gn-o)), (*si-sd-o, Lat. sido),

(*si-skh-o) (for loss of aspiration, see p. 230), (cf. *pet-); (b) athematic: (*si-sta-mi),

, etc. A few verbs have a nasal consonant after the reduplication, e.g. .

3. Nasal stems

(a) (with zero grade of the root): (aor. ), (Doric; Attic is an innovation),

(aor. ). The last is based on a base *kemH2-/kmeH2/ , the -n- being attached to form III of the base

(see PP. 219 f.), cf.

(b) . This suffix is postulated for certain verbs which have a short vowel in Attic and a long vowel in Ionic

(Homer) and so fall into the phonological pattern exemplified by / (see pp. 227 f.): (aor. ),

with an epenthetic vowel: , etc. This suffix was often added to stems with an

infix nasal (of the type Lat. iu-n-g-o, ru-m-p-o, see below): (aor. ), (aor. ),

(aor. ), (Sappho, cf. Latin linquo, Skt. rinakti).

(d)

/ . This type is traceable to a base consisting of a root with a laryngeal extension, a nasal being

inserted between the root and the laryngeal. Thus for the root *dem- 'tame' a base *dem-H2-/dm-eH2-/dm-H2- is

postulated, the second and third forms with infix nasal, yielding Greek

< previous page page_262 next page >

< previous page page_263 next page >

Page 263

/ . Some verbs of this class, under conditions difficult to determine, show an anaptyctic vowel -t-:

(aor. ), , etc. On this vowel, see p. 240.

(e) / . This type had a similar origin to (d): a nasal was infixed in roots extended by -u-. A clear example is

Skt. str-nó-ti (3rd sing.)/str-nu-(u)-ánti (3rd plur.) < *str-n-eu-/str-n-u- (*ster- 'strew', 'cast down'). The expected

Greek correspondents would have the suffix alternation

/ but this appears as / , the change being due

to the analogy of

/ ,: / (the apparent o-grade of the root has been explained as the

reflection of the laryngeal H3, see p. 300). The expected zero grade of the root (characteristic of the II form of the

base) appears in

(*tn-n-u *, *ten- 'stretch'), etc. Verbs with full grade such as ,

etc. are Greek innovations. Note in particular

, which replaced (< *mig-sk-o).

4. . There is evidence that IE roots could be extended by the suffix -s- (see below on the s-aorist, pp.

301 f.), and this could be further extended by a -k-. A good example is Hittite pahs-* 'protect', cf. Latin pas-tor

with the corresponding verb pasco. Besides -s- a suffix -is- also occurs, as in Hittite preterite dai 'put' < *dhe-is.

This likewise occurs in the extended form -isk-, especially in Greek and Armenian, where it is often added to long-

vowel roots. In Hittite -sk- was used to form iteratives while in Latin the main function is inchoative, and this

appears in some Greek representatives (e.g.

). Homer preserves the archaic forms (replaced in Attic

, a -j- formation < gwm-jo*), which is equivalent to Sanskrit gácchati (< *gwm-sk-*), and (*mig-

sk-o). The zero grade is also found in (*pnth-sk-o*; cf. aor. ) with full grade in . From bases

with laryngeal extension we have

(*dhr-eH3-sk-), (aor. , *ml-eH3-sk-), etc. This type

often has reduplication: (*gn-eH3-), (*mn-eH2-. > mna-).

A suffix appears in , and also in and . A suffix , not necessarily identical

with the above, also appears in certain Ionic (Homer and Herodotus) preterites (imperfect and aorist) which have

no augment:

, etc.

< previous page page_263 next page >

< previous page page_264 next page >

Page 264

. The function of the suffix is difficult to determine. Sometimes it seems to express a state arrived at by

action, and perhaps we have here the origin of the aorist passive (see p. 302). Examples of present stems are:

(cf. ), . A suffix -q- also appears in some Homeric aorists:

(*el-/el-eu/el-u-). In and the -q- is added to a stem in . On the

other hand, *mi-n-u- is the infixed form of a base *mei-u-/mi-eu-/mi-u-, and this underlies the Linear B

comparative forms me-u-jo/me-wi-jo 'smaller' (see p. 47).

Stems in -je/o-. Here a distinction is made between a primary use (that is, direct addition to a root) and a

secondary use (attachment to a suffixed root). The suffix was most productive in forming denominatives, especially

those from vowel stems, where it gave rise to the so-called contracted verbs in

, and . Because of Greek

sound developments (loss of inter-vocalic -j- and 'yodization' of consonants (see p. 225)) this, the most productive

of IE present-stem formants, is not directly attested in Greek. In view of the manifold phonetic changes the

formations will be best classified according to the sound (consonant or vowel) immediately preceding -j-.

< *-n-jo *: (< , zero grade of ), (cf. ), (cf. ), ,

etc. This suffix then made itself independent of n-stems and was used to form factitives, particularly (a) the 'colour'

verbs like

, (b) in the verbs 'make hot, cold, dry', etc., e.g. . Other derivatives are intransitive,

denoting physical and mental states, e.g. , etc., such verbs often being appropriately in the middle

voice. A few verbs are primary formations: (cf. aor. , < *mn-j-*).

2. . This formation has close connection with u-adjectives and it appears to have originated in n-extensions like

: , etc. It was later added to o-stems, and the proliferations took place mainly in

three semantic fields: 1, dimensions (

); 2, physical consistency (

), and 3, cosmetic expressions (this perhaps is an extension of 2):

< previous page page_264 next page >

< previous page page_265 next page >

Page 265

, etc. It has been observed that such formations are foreign to pure Attic and are used in particular by Attic

writers whose language is strongly tinged with Ionic.

3. *-rj-, *-lj-:

, etc. (for the phonology, see p. 226).

4. (Attic ). This formant results from the yodization of voiceless dentals and dorsals (pp. 225 f.):

, , etc.

5. . This suffix originated perhaps in the attachment of *-jo to stems in and , but it was subsequently

added to any type of stem, particularly to a-stems: , etc. This brought the verbs into

competition with the type in so that we find encroachments such as , , etc.

. The development of this suffix is parallel with that of , originating in the addition of -jo to (e.g.

, etc.). The compound suffix made itself independent and became attachable to noun-stems of

all types. Its productivity had begun in prehistoric Greek, and it became one of the most prolific of all verbal

suffixes. Such verbs have the meanings (a) to do something, or (b) to be something, connected with the basic noun,

though the precise way in which the noun enters into the activity of the verb varies with each word:

'entertain', 'acquire', 'throw a spear', 'hit with a spear', 'congratulate', 'pronounce happy',

etc.

The denominatives in were originally -j- derivatives from a-stems: , etc. This class

remained relatively unproductive compared with the types in

and . An important group among the

analogical innovations in the classical period comprises the verbs denoting physical ailments (

, etc.)

and abnormal mental states (

, etc.). In the latter sense the compound suffix was especially

prolific:

, etc.

8. . If we disregard the few primary verbs where is due to the loss of a final root consonant (e.g. <

< ) and those from disyllabic bases (e.g. ), this extremely productive type comprises

(a) -j- denominatives from e/o- stems (e.g.

), (b) denominatives from -os/es-

< previous page page_265 next page >

< previous page page_266 next page >

Page 266

stems (e.g. ), and (c) deverbatives in -ejo- with iterative-causative meaning (e.g. ). Although

verbs came to be made from any types of noun-stem, there is a surprisingly small number from simple nouns, this

being the preserve of the suffix -eu- in forming verbs with the meaning 'to be such and such a person', 'to be active

with such and such a thing' (see below).

. These verbs are either factitive in meaning ( 'make clear', 'make blind', etc.) or instrumentative (

'provide with a door', and by extension 'put to death'). A weakened sense of the factitive is present

in verbs meaning 'regard as', 'treat as' (

'regard as just', 'regard with envy', etc.). Since the inherited

denominatives from o-stems are represented by the

type, the class is a Greek innovation, created in analogy

of the

verbs. In the stated functions the class competed with the verbs in and . A number of them

were used in the passive sense 'to be afflicted with a disease', this being also a function of the verbs in

Consequently we find competing formations:

/ , / . The suffix remained productive

in Byzantine Greek and today it is represented by the verbs in

10. . These verbs derive from nouns (pp. 249 f.). The expected form of the suffix < *-ew-jo is attested

dialectic-ally (Elean

), but in general has been restored from the non-present stems and the

underlying nouns. The function of the suffix resembles that of the

class, the sense being 'to be such and such' (

) or 'to be active with regard to such and such' ( ). In the main, simple

nouns form

denominatives, whereas compound nouns favour the type. In Byzantine Greek was much

used in naturalizing Latin loan-words: , (= erogo), etc.

3. Inflexion

A. the Declensional Classes and the Case Inflexions

Greek is an inflexional language in which different monemes are added to the noun stem according to its syntactic

function

< previous page page_266 next page >

< previous page page_267 next page >

Page 267

within the sentence (see above). There may be different sets of such case inflexions with identical functions, as in

the Latin senatus/populus, senatum/populum, senatus/populi, senatui/populo, senatu/populo. The noun stems which

exhibit a given set of case monemes form a noun declension.

Even in IE there are separate forms for all eight cases only in the singular of the o-stems (the thematic declension)

and it is a characteristic of all neuters that they do not distinguish nominative and accusative. In all the other

declensions the ablative and genitive singular are identical. In the plural IE had only six case distinctions: the

dative and ablative are identical and there is no vocative form. In the separate languages there were further

simplifications. Certain fossil survivals apart, Greek combined the genitive and ablative, and the dative with the

locative and perhaps the instrumental. This phenomenon is known as the 'syncretism' of the cases. The following

table sets forth the case endings postulated by comparatists for IE:

Singular

Dual Plural

Nominative -s,

-e, -es

Vocative

-e, -es

Accusative -m/

-e, -ns/-ns *

Genitive -es/-os/-s -ous -om/-om

Ablative -es /-os /-s, -ed/-od -bh(y) o, -mo -bh (y) os, -mos

Dative -ei -bh (y) o, -mo -bh (y) os, -mos

Instrumental -e/-o, -bhmi/-mi -bh(y) o, -mo -bhmis/-mis, -ois

Locative -i -ou -su

stands for zero inflexion, which in certain declensions of the animate gender is accompanied by extension

(vrddhi) of the suffix. On the distribution of the alternative forms (e.g. the m-cases) in the IE dialects, see p. 6.

The case monemes show Ablaut alternation: e.g. genitive -es/-os/-s, ablative -ed/-od,1 instrumental -e/-o. The same

is true of the stem, the general rule being that the full grade appears in the nominative, vocative and locative of the

singular and the nominative, vocative and accusative of the plural, these being the 'strong' cases. In the nominative

singular of certain classes an extended grade of the Ablaut appears. The other cases have the zero grade of the

stem. This ideal picture was, of course, obscured by frequent analogical phenomena.

1 The long vowels result from the contraction of -ed and -od with the preceding thematic vowel.

< previous page page_267 next page >

< previous page page_268 next page >

Page 268

Even in IE contraction took place when a case beginning with a vowel was added to a vowel stem and this was the

beginning of the process which diversified a single original IE declension into the different declensional classes

presented by the descendant languages. These are best organized and analysed according to the final sounds of the

stems. The thematic declension (o-stems) stands apart from the athematic group, comprising stems in -a, plosives,

liquids and nasals, s-stems, and stems in -i, -u and diphthongs.

1. O-Stems

Singular

Nom. *-s: . Neuters do not distinguish nom. and acc., both of which have the ending -om.

Voc. bare stem with e-grade: .

Acc. *-m, which regularly > -n in Greek: .

Gen. *osjo:

Dat. *oi (< *-o- + *ei): .

Dual Nom., Voc., Acc. *-o ( < *-o + *-e since -e is the ending in the consonant stems.

Gen., Dat. *-ous.

The Greek has no parallel elsewhere, but Arcadian has the form . The Sanskrit -ayos is thought to be

an extension (the -ay- originating in the pronouns and in the word 'two', dvayos) of the original ending *-ous, to

which the O.Sl. ending is traceable. In view of the Mycenaean locative form duwoupi 'in two (land portions)' it is

possible that the Greek forms in

also owe their genesis to the influence of the word for 'two', though the

spelling rules make it impossible to decide whether the Linear B word stands for dwounphi or dwouphi. It is just

conceivable that the Greek ending went through the stages

, Attic having further the dissimilated

form

for .

Plural

Nominative *-os < *-o-+ *-es. This was replaced in Greek as early as the Mycenaean period by -oi taken from the

pronominal declension (see below). The neuters utilized as their plural an old collective formation (still taking the

singular verb in Attic

< previous page page_268 next page >