Gonda Periphrastic Forms

A Remark on 'Periphrastic' Constructions in GreekAuthor(s): J. GondaSource: Mnemosyne, Fourth Series, Vol. 12, Fasc. 2 (1959), pp. 97-112Published by: BRILLStable URL: http://www.jstor.org/stable/4428220 .

Accessed: 03/03/2014 04:28

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

BRILL is collaborating with JSTOR to digitize, preserve and extend access to Mnemosyne.

http://www.jstor.org

This content downloaded from 212.189.140.54 on Mon, 3 Mar 2014 04:28:07 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=bap

http://www.jstor.org/stable/4428220?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


A REMARK ON 'PERIPHRASTIC CONSTRUCTIONS

IN GREEK

J. GONDA

In a recent article *) Prof. H. B. Rosen of Jerusalem drew

attention to the Greek constructions of the type ???? . . . ?st?, ???? . . . ??, ???? . . . e??a? beside ??e?, ???e, ??e??. Basing himself

especially on a complete research of the relevant passages in Hero-

dotus he arrived at the conclusion that these periphrastic con-

structions were, by this historian and other ancient Greek authors, used when the idea expressed by the verb is not, logically speaking, the predicate of the sentence, but the subject, and when at the same

time the 'logical' predicate is not the agent of the verb, but a

complement to the latter, the term complement being taken in a

large sense including inter alia, object, an adjunct, etc.

Rosen's argument, however interesting, could, in my opinion, be

clarified if this conclusion were restated as follows2). A sentence

such as Hdt. IX 15, 4 (16) ?? d? t? de?p??? p??e??e??? ?? T???s? does

not exactly mean "the meal took place at Thebes", but rather "it

was at Thebes that the meal took place", and similarly I 146, 3 ta?ta d? ?? ?????e?a ?? ????t?? "it was at Milete that those things came to pass". That is to say, in translating these sentences into

English (or French and other languages) a "cleft sentence" (phrase

coup?e) must as a rule be preferred if the sense is to be rendered

as exactly as possible. Even K?hner-Gerth 3) who whilst giving a

mere enumeration of periphrastic structures did not notice this

point, translated the former passages: "der Ort, wo das Mahl ver-

anstaltet wurde, war Theben". This means that it is the contents of

the verb which is the thema of the sentence, not the subject proper from the logical point of view. The idea contained in the verb is the

thema i.e. starting-point of the speaker's or author's argument or

?) H. B. Ros?n, Mus. Helv. 1957, x33 ff?

2) I limit my observations to the type of sentence discussed by Rosen.

3) R. K?hner-B. Gerth, Ausf?hrliche Grammatik der griechischen Sprache*, I. 38, ?. 3.

Mnemosyne, XII



98 A REMARK ON 'PERIPHRASTIC' CONSTRUCTIONS IN GREEK

communication. The adjuncts (or objects etc., in general 'the

complements') are the propos, that is, that which the speaker or

author wants to bring forward with regard to the thema, 'the

predication'. In the first sentence the taking place of the meal is

the thema, the fact that this event took place at Thebes is the

propos. The context had made it clear that a dinner would be held.

The new fact which is brought to the reader's notice with, perhaps, some emphasis is that the place where it took place is Thebes. This

frequent construction is often found in sentences expressing an

'antithesis' : II 134, 2 ?at? "??as?? ?as??e???ta ?? a??????sa

'??d?p??, a??' ?? ?at? t??t??, in contradistinction to the mere

statement of a historical fact in III 57, ? ta d? t?? S?f???? p????ata

???a?e t??t?? t?? ??????. Not rarely ??? ... d? . . . are added to

underline the contrast: II 99,1 ; VII 3, 3; or the idea of opposition is expressed by antithetic pronouns (e.g. I 112, 3) or ????? (e.g. II

48,2). In the same way, Soph. O.R. 968 ff. wrote e?? d' dd' ????de /

??a?st?? e????? - e? t? ?? t???? p???? / ?at?f???' ? ??t? d' a? ?a???

e?? '? ????, words rendered by Storr1) as follows: "and here am I

who ne'er unsheathed a sword ; unless the longing for his absent son

killed him and so / slew him in a sense". It is the words ?? e??? which are thrown into relief, also by the position of emphasis in

which they appear. It is well known that in spite of the general

principle by which emphatic words tend to an initial, or at least

early, position they are, in certain cases, also placed at the end of

a clause or sentence 2). K?hner-Gerth made the attempt to express this nuance by interpreting "dann w?re er freilich ein von mir

get?teter" ("dann w?re ich sein M?rder"). It is clear that the succinct treatment of these constructions by

Schwyzer-Debrunner 3)?who say that they are "eine expressive

Umschreibung des verbum finitum"?is not complete and exhaust-

ive. K?hner and Gerth's statement4): "Soll der Verbalbegriff

?) Sophocles, with an English translation by G. F. Storr (Loeb), London- Cambridge, Mass., 1956.

2) For particulars see J. D. Denniston, Greek Prose Style (Oxford, 1952), 45 ff-

3) E. Schwyzer-A. Debrunner, Griechische Grammatik, I (M?nchen, 1939), 812, sub a 4; II (M?nchen, 1950), 407 f.

4) K?hner-Gerth, I.e.



A REMARK ON 'PERIPHRASTIC CONSTRUCTIONS IN GREEK 99

selbst?ndiger und nachdr?cklicher hervorgehoben werden, so tritt

an die Stelle des einfachen Pr?dikatsverbums eine Umschreibung durch das Partizip Pr?sentis, Perfekti oder Aoristi (letzteres fast

nur dichterisch) mit der Kopula e??a?" is highly inaccurate, if not

incorrect : it is not the idea expressed by the verb which is thrown

into relief but the complements added to it. In Soph. O.R. 90 there is a slight antithesis between the preceding sentence est?? d?

p???? t??p?? which may imply uncertainty or even?notwithstand-

ing the reassuring words of Creon in 87 f.-?apprehension, and the

last part of 89 and 90 ??te ??? ??as?? ??t' ?d? p??de?sa? e??? t?? ?e ??? ????? "thus far thy words give me no ground for confidence or

fear". In Hdt. II 10, 3 it is not the 'Verbalbegriff' ?p?de???e??? that is emphasized: e?s? d? ?a? ????? p?ta???, ?? ?at? t?? ?e????

???te? ?e???ea, ??t??e? ???a ?p?de???e??? ?e???a e?s?. Another

attempt to formulate the difference between the synthetic and the

analytic construction was made by Rehdantz and other commen-

tators: "die Umschreibung des Verb, finit, durch e??a? mit dem

Partizip ist eine Abl?sung der Kopula von dem Pr?dikatsbegriff, durch welche beides selbst?ndig hingestellt und der Ausdruck

gewichtiger wird" x). This definition would for instance apply to

Xen. Anab. II 2, 13; III 3, 2; IV 3, 5; IV 5, 15. In Anab. IV 1, 3 the words ?a? est?? ??t?? ???? are according to the same authorities

"st?rker als ??t?? ??e?". The question may however arise why this

"gewichtiger Ausdruck" should have been preferred and what it

was intended to express.

What, however, seems to be essential is that the author prefers a semi-nominal construction to a verbal sentence. Whereas from

prehistoric times verbal sentences were mainly used in narratives

and accounts of successions of facts, drawing attention to the ever

changing events and occurrences described, nominal or semi-

nominal sentences were largely preferred in descriptions, statements,

elucidations, explications, characterizations, exclamations, in-

dications of time or circumstances, transitional formulas etc. 2)

?) Thus C. Rehdantz and O. Carnuth, Xenophons Anabasis, II6 (Berlin, 1905), 17.

2) See e.g. A. Meillet, La phrase nominale en indo-europ?en, M?m. Soc.

Ling. 14, iff.; B. Delbr?ck, Vergleichende Syntax, III (Strasburg, 1900), III, ii7ff. ;H. Hirt, Indogermanische Grammatik, VII (Heidelberg, 1937), 19 ff.



100 A REMARK ON PERIPHRASTIC CONSTRUCTIONS IN GREEK

in which the attention of the hearer is drawn to the static nominal

predicate. Whereas the predicate of a verbal sentence focusses

the hearer's thoughts on what happens, the predicate of a nominal

sentence directs his attention to what is. This is especially evident

in combinations of participial groups and adjectives: ? 6? ?e??a???? t ?s??es?a ?a? ?? deda???te? ????? ("knowing nothing of, nichts

verstehend von") 1). "Es wird durch den nominalen Stil der Vor-

gang, das Ereignis des Geschehens aus der zeitlichen Atmosph?re in die r?umliche verschoben" 2). For instance in ? 488 where a

semi-nominal participial construction is preferred : ????a? d' ?? t???

f??? pef??????? e??e?a? a?d???, it is the situation or state of affairs

which is thrown into relief rather than an event or occurrence.

Being more graphic in character and distracting the attention of

the audience from an individual point of time at which something narrated happens, the semi-nominal expression is apt, on the one

hand, to create the impression of being more forceful?compare in

English : who was she to be refusing what might prove to be, perhaps, the last request of her dying husband??and on the other hand, to

bring out the object or adjuncts amplifying the predicate. The

transition from the 'narrative verbal style' to the 'descriptive semi-

nominal style' may be illustrated by Xen. Anab. IV 5, 15 d?a ta?

t??a?ta? ?d? ?????a? ?pe?e?p??t? t??e? t?? st?at??t?? ?a? ?d??te?

???a? t? ??????, d?a t? ???e???p??a? a?t??? t?? ????a, ??a??? tet????a?? ?a? ?tet??e? d?a ?????? t??? ? p??s??? ?? ?t?????sa ?? ??p??. ??ta??'

??t?ap??e??? ??????t? . . . and IV 3> 5 a^ d? ???a? a?ta?, ef' ??

pa?ateta?????? ??t?? ?sa? t??a ? t?tta?a p????a ?p? t?? p?ta???

?pe???? ?d?? d? ??a ??????? ?? ????sa ??? ?spe? ?e???p???t???

ta?t? ? ?pe????t? d?a?a??e?? ?? '?????e?. With regard to ? 269 ??d?

?a? ??d' ??? f??? ?e?as????? ?'??e?a? a???? "for I deem that I too

am not forgetful of valour" Ameis-Hentze observed that "die

Umschreibung die Charactereigenschaft bezeichnet". It would

perhaps be somewhat more to the point to say that a static idea of

being forgetful of valour is, in this verse, negated rather than a

?) See G. Bj?rck, ?? ????S?O?. Die periphrastischen Konstruktionen im Griechischen, Skr. Vet. Samf. 32, 2 (Uppsala, 1940), 17 ff.; 32.

2) M. Deutschbein, Neuenglische Stilistik (Leipzig, 1932), 140.



A REMARK ON PERIPHRASTIC CONSTRUCTIONS IN GREEK IOI

dynamic process of forgetting valour1). Hence also the sudden

transition from a finite form in a relation of a succession of events

in a 'narrative' passage to a periphrasis in a reference to a state of

affairs: Hdt. V 77, 3 t?? d? p?da? a?t?? . . . ??e????asa? ?? t??

????p???? a? pe? et? ?a? ?? ??? ?sa? pe??e??sa?. The difference between periphrasis and the finite verb is especially

obvious when both constructions occur in the same context: in

Hdt. II 95, I the words p??? d? t??? ????pa? . . . t?de sf? ?st? ?e-

???a?????a are purely 'descriptive' stating that the Egyptians have definite contrivances against the gnats, but the sentence

t??s? d? pe?? ta e?ea ??????s? t?de . . . ?e??????ta? means, in a

'narrative' style of expression, that some people had contrived

some defence. Compare also VI 65, 2. It may be of some interest

to make a comparison between Hdt. Ill 133, 2 . . . de?ses?a? d?

??de??? t?? dsa ?? a?s????? ?st? f????ta and ? ??, 2 . . . ?a? ??d?a

?f???a? ?????? ?? a?s????? ?e????? f??e?. A transition from a narra-

tive of 'doing' to a statement of 'having' involves a change of con-

struction in Hdt. II 37, 4 ??te t? ?a? t?? ??????? t?????s? ??te dapa-

???ta?, ???a ?a? s?t?a sf? ?st? ??? pess??e?a ... "... they have

bread viz. baked grain grown on the sacred domains". It might

easily be imagined that the words ?. p. are a stereotyped ampli- fication or 'epexegesis'. References to a state of affairs, to a per-

manency, or a static representation of the process expressed by the verb occur e.g. also Hdt. IX 27, 5 ???? d? e? ??d?? ???? ?st?

?p?dede?????? . . ., a??? ?a? ?p? t?? ?? ?a?a???? ????? ????? e??e?.

Compare e.g. also Soph. Tr. 446. The similarity to semi-nominal

adjective constructions and the semi-nominal character of the

periphrasis is also clear in instances such as Hdt. I 45, ? ????? t?? te

p??t???? ???t?? s??f????, ?a? ?? ?p' ??e???? t?? ?a???a?ta ?p????e???

e??, ??d? ?? e?? ???s????. Participles may indeed more or less in-

cidentally fulfil the function of adjectives: Hdt. Ill 39, 3 . . . t??

???????te?? t? p????ata a??et? ?a? ?? ?e?????a ??? te t?? ?????? . . ., where ?. is pe?????ta "celebrated".

The retardative force of the periphrastic form is also clear in the

?) The Homeric instances seem to have escaped the attention of P. Chantraine, Grammaire hom?rique, II (Paris, 1953), 201, who mentions only ? 257.




sentence type Hdt. I 112, 3 ?a? ??t? ??te s? ???sea? ?d????? . . .

??te ???? ?a??? ?e????e????a esta?? d te ??? te??e?? ?as??????taf??

????se? ?a? ? pe??e?? ??? ?p???e? t?? ?????, the clause ??te ?. ?. ?. e.

being explained by the next sentence. In IV 128, 2 t? ?e????e????a is equivalent to t? ????e??ata. Another case of periphrasis by means

of a participle which is not only in frequent use, but also employed as a substantive occurs II 170, 2 ????? te ?st? ??????? .... ; cf. e.g. I 134 ?? ????e??? "the neighbouring people". Iri a similar way beside Hdt. VII in, 2 ??t?? t? ?a?t???? e?s? ?e?t?????? we find ?

?e?t?????? "the owner". The same retardative force of the con-

struction under discussion is also obvious in cases such as Hdt.

II i55> 3 where a description o? a sanctuary is interrupted by the

words t? d? ??? t?? fa?e??? ?? ????a ????st?? pa?e???e??? f??s?. It seems worth noticing that translators whilst resorting to a

cleft sentence to render the meaning of the original Greek as

exactly as possible preferred also a nominal equivalent of the Greek

participle to a French or English verb: Hdt. I 86, 3 t?? d? ????- s?? . . . ?se??e?? . . . t? t?? S??????, ?? ?? e?? s?? ?e?? e????????, t? ... . : "that there was a divine warning in the words . . ."

(Rawlinson) ; "qu'il y avait une inspiration divine dans ce mot que S. lui avait dit" (Legrand).

That the sudden transition to a state of rest, the arrest of the

course of events narrated is, however, a suitable means of bringing other elements of the sentence to the fore may appear from Xen.

Anab. Ill, 3, 2 ???st?p????????? d? a?t?? ?'??eta? M. . . . ?a? . . .

???e? ?de* ??? . . . ?a? ????? p?st?? ?? . . . ?a? ??? ???? e?????? ?a?

????de d' e??? s?? p????? f???? d?????. An instructive comparison

may be made between E 383 ff. and E 873. In E 383 a statement

is made about a fact, the hearer's attention is focussed on a process: p????? ??? d? t???e? . . . / ?? a?d???. In substantiation of this

statement examples are adduced and the form t???the aorist

emphasizing the mere verbal idea?is three times in succession

anaphorically repeated: 385 t?? ??? "???? ?te ... "so suffered Ares,

who...", 392, and 395 t?? d' ??d??.. . ???? ??st??, drawing special attention to the processes. In E 873, on the other hand, the peri-

phrastic construction describes a situation or state of affairs extant

at the moment of speaking?"fasst alle einzelnen F?lle bis zur



A REMARK ON PERIPHRASTIC CONSTRUCTIONS IN GREEK IO3

Gegenwart zusammen" x)?, but in so doing it does not divert the

attention from the words amplifying the verbal phrase, that is to

say from the object ????sta which comes already to the fore by its

very position and the opposition in which it is to ????? in 874. It is of course true that a 18 ??d' ???a pef??????? ?e? ?????? / ?a?

?et? ??s? f????s? a situation is described, and the course of succes-

sive events referred to by the preceding verbs f????, ?sa?, e???e,

???e is arrested, but it is the idea expressed by ??d' ???a reinforced

by its initial position and by the anaphorically added explanation which is thrown into relief. In ? 455 the pith of the matter is from

the psychological point of view the idea expressed by ?? p?, not

that of escaping or being in safety, whatever other motives there

were for preferring a static expression to a dynamic verbal sentence :

?dt??, d? ?? p? f??? pef??????? e??a? ??e????. Some other examples may be subjoined here: Soph. At. 1324

????se? a?s???? d??? ?a? ?? t??a?ta ?e "I had reviled him; for vile

were his deeds toward me" 2) ; Plato Soph. 217C ?? t?????, ? ???e,

???? t?? ?e p??t?? a?t?s??t?? ????? apa????e?? ?????, t?s??de d' ????

f???e "veuille donc, ?tranger, ? la premi?re faveur que nous te

demandons, ne point opposer de refus. Mais plut?t, dis-nous" (Di?s). There remains, however, another remark to be made. As is well

known, it is in translating Latin or Sanskrit participles sometimes

necessary to substitute infinitives or nouns of action: Plaut. Epid.

144 ante solem occasum; Bhagavadg?t? 3, 35 ?reyan svadharmo

vigunah / paradharm?t svanusthit?t "it is better (to perform) one's

own duty (norm) imperfectly than to perform the duty of another

perfectly"; Pa?catantra, Intr. 3 kanyd. . . janit? "the birth of a

daughter" 3). That means that the nucleus or central element of the

word group is contained in the participle. Passages such as the

above Hdt. IX 15, 4 which retard the progress of the narrative and

in which the initial e??a? may be taken to have meant, or to have

developed from, "there is, it happens", admit of the same inter-

?) ?. F. Ameis-C. Hentze, Homers Iliass (Leipzig t Berlin, 1927), 100.

2) Cf. R. C. Jebb, Sophocles, VII (Cambridge, 1907), 197. 3) For other examples see J. S. Speyer, Sanskrit Syntax (Leyden, 1886),

2Q2 f. ; the same, Vedische und Sanskrit-Syntax (Strassburg, 1896), 97. Latin syntax has exerted its influence upon, for instance, Modern German; see O. Behaghel, Deutsche Syntax, II (Heidelberg, 1924), 408 f.




pretation1). Thus Xen. Anab. II 2, 13 ?? d? a?t? ? st?at???a ??d??

???? d??a???? ? ?p?d???a? ? ?p?f??e?? may be translated by: "the

signification of that piece of strategy was nothing else but running

away . . ." ("es bedeutete diese Heeresf?hrung . . . "2)). Compare also instances such as Hdt. I 57, 1 f. ??t??a d? ???ssa? ?esa? (narrative style, event, process) ?? ?e?as???, ??? ??? ?t?e???? e?pa? . . .

(but if we may judge by some data we possess) ?sa? ?? ?e?as???

????a??? ???ssa? ???te? "... then it was a foreign language that

the P. spoke", or ". . . then the language of the P. was a foreign one" 3) ; Thuc. I 99, 2 4).

Now the question arises whether Rosen's view of these con-

structions is in harmony with the line of thought developed in

connection with periphrasis in Greek by Bj?rck 5). In a publication which in spite of its title is chiefly concerned with the syntax of the

New Testament and other texts of the later period, and which does

not appear to have been noticed by Rosen, the Swedish scholar

while sharply distinguishing between "improper periphrasis"?in which the participle is no predicate, or e??a? no copula?, adjectival

periphrasis?in which the participle has the character of an ad-

jective?, and real periphrasis?which is the subject of his book-

makes an attempt to show that the type e??a? + present participle

may be fairly considered the Greek counterpart of the well-known

English "progressive tenses" 6). Thus NT. Ev. Luc. 19, 47 ?a? ?? d?d?s??? t? ?a?' ????a? ?? t?? ?e???? ?? d? a???e?e?? ?a? ?? ??a??ate??

???t??? a?t?? ?p???sa? ... is the syntactic counterpart of the

English: "and he was teaching daily in the temple. But the chief

priests and the scribes . . . sought to destroy him" in that ?? d?d?s??? as well as he was teaching characterises the process

(action or state) as being conceived as a temporal frame encom-

?) See J. Gonda, Remarques sur la place du verbe . . . (Utrecht, 1952), 67 ff.

2) C. Rehdantz-O. Carnuth, Xenophons Anabasis I7 (Berlin, 1912), 133. 3) D. Barbelenet, De la phrase ? verbe ?tre dans Vionien d?H?rodote (Thesis

Paris, 1913), 92, has missed the point in limiting himself to the remark that the sentence contains a commentary.

4) These cases should be distinguished from the constructions discussed by Bj?rck, o.e., 14 (Arist. Eq. 225 etc.).

5) G. Bj?rck, ?? ????S?O?. 6) See especially Bj?rck, o.e., 41 ff.



A REMARK ON 'PERIPHRASTIC CONSTRUCTIONS IN GREEK I05

passing something else which is often expressed in the context, but

may also be understood from the whole contextx). The purport of

these forms is, according to Bj?rck, not to express duration in it-

self, but relative duration, compared with the shorter time occupied

by some other process. There is, however, room for the observation that Bj?rck has, on

the one hand, too onesidedly focussed his attention on this function

of the English construction which, however important, is not the

only one, and on the other hand has laid too much emphasis on

the above character of the type ?? d?d?s???, while drawing, some-

what inconsiderately, conclusions as to the classical period from an

examination of the later facts. Whereas Rosen has paid no attention

whatever to the function and the particulars observed by Bj?rck, the latter has, like many other scholars, entirely missed the con-

siderable element of correctness in the argument of the former.

There are indeed ancient examples, even in Herodotus, of the

phenomenon discussed by Bj?rck. Hdt. VIII 137, 4 ???a?ta ?

?as??e?? t?? ??s??? p??? ????sa? ? ?? ?a? ?at? t?? ?ap??d???? ?? t??

????? ?s???? ? ????? ? e?pe . . . ??s??? de ???? ??? ????? ????? t??de

?p?d?d???, de??a? t?? ?????: ". . . now it happened that the sun was

shining down the chimney into the room" (Rawlinson). Bj?rck 2) is no doubt right in stating that "die progressive Form ausge- zeichnet f?r die parenthetisch eingef?hrte 'Rahmenhandlung'

passt". Cf. also I 152,1; VI 103,4 (t????a?ta); Soph. Tr. 578 f.

t??t' ?????sas', ? f??a?, ? d????? ?a? ?? / . . . ???e???????? ?a???,

? / ??t??a t??d' ??a?a. That this phenomenon is not foreign to

other authors either may appear from Thuc. VII 50, 4 ?a? ? ????a? ?

?? ??? t? ?a? ??a? ?eas??? te ?a? t?? t????t?? p??s?e??e??? ? ??d'

a? d?a????e?sas?a? et? ef?, p??? . . ., where Marchant observes that

"the tense of e??? must precede the participle in this periphrasis, as it is emphatic, representing a state of things existing at the

time referred to" 3). There is on the other hand no denying that in cases such as Hdt.

? ) See especially O. Jespersen, A Modern English Grammar, IV (Heidel- berg, 1931), 164 ff., and especially, 180; the same, The Philosophy of Gram- mar (London, 1935), 277 ^?

2) Bj?rck, o.e., 71. 3) E. C. Marchant, Thucydides, VII (London, 1952), 171.



??6 A REMARK ON 'PERIPHRASTIC' CONSTRUCTIONS IN GREEK

VIII 137, 4; I 152, I the adjuncts (?? t?? ?????; ?at? . . . t????) are

to receive some special attention. Cases are however not wanting in which the clause or sentence containing the construction under

consideration does not, properly speaking, constitute a parenthesis: Thuc. I 47, 3 . . . ?a? a? ?tt??a? d??a pa??sa? . . . ?sa? de ?a? . . .

p????? t?? ?a?????? pa?a?e???????te? ; I gg, 2; II 12, 2.

There are however exceptions to the rule that the periphrastic construction represents a process as a temporal frame encompassing another event : Hdt. ??3?,2?d? ??t?? a?t? ? t?? ??f?p???? ... ap?-

?a?e t?? ?e??a?, ?a? ??? t?? e????a? a?t??? e??a? pep?????a? t? pe? a?

??a? ?pa???. Whereas in the above sentence the periphrastic phrase refers to a present state of affairs, it is elsewhere used to indicate

continuance in the (relative) past: I 103, 1 . . . p??t?? te ?????se ?at? t??ea t??? ?? t?? ?s??? ?a? p??t?? d??ta?e ????? e??st??? e??a?

. . . ? p?? t?? d? ??a??? ?? p??ta ?????? ??apef?????a.

Another passage to which the explication proposed by Bj?rck does not apply is Hdt. II 134, 2 (see above) ; II 99, ? ????? ?e? t??t?? ???? te ??? ?a? ????? ?a? ?st???? ta?ta ?????sa ?st?, t? d? ap?

t??de ????pt???? ?????a? ?????? ????? "thus far I have spoken of

Egypt from my own observation . . ." (Rawlinson), or "thus far

it is my own observation on which the story is based . . .". Other

passages might indeed be adduced by both scholars in favour of

their theories. Thus Hdt. VII 190 ???' ? ??? t???a ??? e?t?????

e????as? ???a p???s??? ????et?? ?? ??? t?? ?a? t??t?? ??a??? s??f??? ??pe?sa pa?d?f???? *). Here also, the present author would however

point out the nominal ? and hence 'static5 ? character of the

second part of the communication from which the phenomenon detected by Rosen as well as the character of the sentence described

by Bj?rck seem to be easily explainable. See also Hdt. IV 78, 3 ?

S????? d?a?t?? ??? ??da??? ???s?et? S?????? ?, ???a p????? p??? ta

???????? ?????? tet?a?????? ?? . . . ?p??e? te t????t?, and in addi-

tion: I 102, 2; I 146, 3; I 175, 1; V 47, 7; IX 49, 3. In a description of the valuable products of the far-off countries Herodotus III 107, 1

after three times ?st?, f??e?, ?????ta? and another ?st? wrote: ?? d?

?) "Denn kein Mensch geniesst eines ungetr?bten Gl?ckes" (H. Stein, Herodotus IV, Berlin, 1861, 177); "der hervorgehobene Satzteil mit ?a? eingeleitet5* (Rosen, o.e., 144).



A REMARK ON PERIPHRASTIC CONSTRUCTIONS IN GREEK IO7

ta?t?? ???a??t?? t? ?st? ?????? ?????? pas??? f???????? ?a? s????? ?a? ?as?? . . ., continuing: ta?ta p??ta p??? t?? s?????? d?spet??? ?t???ta? ?? %???????. The retardative force of the 'descriptive* semi-

nominal construction indicating that the process continues at the

moment of writing helps to throw ?????? into relief (Rosen) as well

as to denote the larger compass of time within which the process ?t???ta? takes place (Bj?rck).

Besides being one-sided Rosen's explication has, if I am not

mistaken, sometimes overreached itself. Too often he has given the

impression of one who is too subtle in attempting to show that the

construction was chosen in order to emphasize one of the elements

of the sentence other than subject and predicate. A type worth

mentioning in this connection is Hdt. V 1, 3 ??? a? e?? ? ???s??? ?p?te?e? ????? ????, ??? ???te??? t? ?'????. I must confess that there is

something to be said for Stein's comment: "?p. im Begriffe sich

zu erf?llen" appears to be more convincing than the view that the

construction chosen is due to the emphasis laid on the pronoun. It is for instance difficult to believe that in Hdt. VII 65 it is ??t?? which conditions the occurrence of a periphrasis: the sentence

?sta?????? ??? d? ?sa? ??t?? ??d?? is a transitional formula which

brings a descriptive passage to an end; the following finite verb

p??setet??at? resumes the course of the events which are the

subject of the narrative. Rosen's explication of VI 65, 2 *) is right as far as it goes, but the motive inducing the author to prefer the

periphrastic construction was not the necessity to emphasize the

words d?a p????a t????de, but the 'parenthetical' character of the

sentence which whilst interrupting the course of the narrative

informs the reader of a past event explaining a state of affairs.

Among the examples quoted by Rosen is also Hdt. Ill 28, ? e'? ?e??

t?? ?e??????? ?p??????? e?? ????pt???s? "si c'?tait un dieu traitable

qui ?tait venu aux Egyptiens" which exhibits a verbal form which

he elsewhere excluded from the 'second tenses', 2) because it does

not occur beside a synthetic form. This passage however shows that

the phenomenon vindicated by him is not limited to 'second

tenses'.

i) Rosen, o.e., 145. 2) Rosen, 147; 137.



??8 A REMARK ON 'PERIPHRASTIC' CONSTRUCTIONS IN GREEK

As neither author has dealt with the 'origin' of this construction

some further observations bearing upon this point may perhaps find

a place here. Passages such as Hdt. Ill 105, 2 are of special interest

because they may perhaps throw some light on the history of this

construction. After having told the famous story of the gold-

digging ants, Herodotus observes: t?? ??? d? p??? t?? ???s?? ??t?

?? '??d?? ?t???ta?, ?? ???sa? fas?? ????? d? spa???te??? ?st? ?? t??

????? ???ss??e???. There is?it is true?an antithesis and, hence, some 'emphasis' laid on the adjunct. Legrand *) is however no doubt

right in suggesting ?st? to have been a complete verb and ???ss?-

?e??? an appositive participle?or rather: in suggesting that this

type of sentence may have developed from ?st? "il y a" etc. ? : "ils

en ont d'autre, en moindre quantit?, qu'ils extraient des mines".

Hdt. VII 179 while exemplifying the tendency discussed by Bj?rck

may be considered to represent, or to go back to, an appositional

type of expression: ? d? ?a?t???? ????e? st?at?? . . . pa???a?e ???s?

t??s? ???sta p?e??s??s? d??a ??? S??????, ???a ?sa? p??f???ss??sa?

??e? t?e?? ??????de? . . . There can be hardly any objection to a

translation: "... where were, being on guard, three Greek vessels

. . ."; compare also Legrand's rendering: "l? se trouvaient de garde en avant-poste trois vaisseaux grecs". It is on the other hand true

that, ???a being the starting-point of the clause introduced by that

word, ??e? t. ?. are the propos which is thrown into some relief; and that there is also room for the contention that the process

expressed in this clause was continuing at the time when the acti-

vity denoted by pa???a?e occurred.

Although Rosen is, from his point of view, right in regarding Hdt. VII 2, 2 ?sa? ?a? ?a?e??? ?a? p??te??? ? ?as??e?sa? ?e????te?

t?e?? pa?de? ?? t?? p??te??? ???a???? ... as an instance of a 'Gegen- satzkonstruktion' it is possible to take this passage as an exact

counterpart of the Dutch "er waren D., al voordat hij koning werd, drie zoons geboren uit zijn eerste vrouw" : if ?e?. ?? t. p. ?. is con-

sidered an apposition?cf. e.g. Thuc. II 67, ? ?? ?? st??te??a t??

'????a??? p????????? ? ?sa? is the initial "il y avait". It must

in this connection be borne in mind that in a considerable number

of the instances quoted by Rosen the form of e??a? occupies the

?) H?rodote (coll. Bud?).



A REMARK ON PERIPHRASTIC CONSTRUCTIONS IN GREEK I09

initial position; cf. e.g. IV 165, 2; VII 190. The sense of "il y a; there is" is still clearly perceptible in cases such as I 210 as well:

?? e?? ???? ???s?? ?e?????, dst?? t?? ?p?????e?se?e "... that there

should be a Persian living" (Rawlinson). In accordance with

Legrand's felicitous translation the words Hdt. I 160, 5 ?? d? ??????

??t?? ??? ?????? ?e???e???, dte . . . mean "et il y eut un temps, un

temps qui fut assez long ..." : i.e. ?? is "es war" x). We may indeed easily imagine that sentences of the type Hdt.

Ill 60, 1 reflect an older construction with the verbum existentiae:

??????a d? pe?? Sa???? ??????, dt? sf? t??a ?st? ????sta ap??t??

??????? ??e??as???a: Legrand's translation "parce que c'est chez

eux qu'ont ?t? ex?cut?s les trois ouvrages les plus grands" might,

then, be slightly modified. And why should I 51, 1 . . . ?? t?? ??s?? ???a? ?

? d? ?st? ?p? t?? '?s?p?? . . . d??a stad???? ap????sa not have

originated in: "and that (island) was (away) from 2) the ?., (lying) at a distance of ten s." ?

It may be remembered that to indicate a process which extends

over a period of time that began in the past and includes the pre-

sent, and usually to indicate that the process referred to may continue in future the English language uses the present perfect

progressive tense: it has been raining since early morning. The

sentence he has been taking violin lessons this year implies that he

continues to take these lessons. An example of a Greek construction

with a perfect participle is Hdt. IV 22, 2 s??e??e? d? t??t??s? . . .

?at????????? e?s? t??s? ?????a ?e?ta? ????a?: the implication is that

the state described still continues.

In a considerable frequency of cases Herodotus combines a form

of e??a? and a perfect participle, almost always to express a state or

situation?resulting from a previous process?continuing to occur

at the moment when some event or other is being narrated to occur,

and hence representing that situation as being, to a certain extent,

in opposition to, or different from, other occurrences. Compare e.g. VI 44, 1... t?? pe??? ?a?ed??a? p??? t??s? ?p?????s? d?????? p??s-

e?t?sa?t?? ta ??? e?t?? ?a?ed???? ?'??ea p??ta sf? ?d? ?? ?p??e???a

?) Cf. also Stein, o.e., I, 131. 2) For ?p? compare e.g. ? 292; ? 53; Thuc. I 7; I 99, 3, and especially,

Xen. H. G. II 4, 4.



HO A REMARK ON 'PERIPHRASTIC' CONSTRUCTIONS IN GREEK

?e????ta ". . . car en de?? de la M. tous les peuples ?taient d?j? rang?s sous leur domination" (Legrand). Similarly VI 33, ?. To indicate that a state was still continuing at a point or period of past time referred to in the context a perfect participle may be connected with a past finite form of e??a?: Hdt. VII 212, 2 . . . s????a????. o? d? '?????e? ?at? t???? te ?a? ?at? ?'??ea ?e??s??????? ?sa?, ?a? ??

???e? ??ast?? ??????t?, a construction suitable to direct the hearer's attention to the words ?at? . . . e??ea. There is therefore room for the assumption that the tendency to avoid the old form in -ata?, -at? etc. x) has not been the only factor in the process which led to a greater use of periphrastic forms 2).

Among the examples adduced by Rosen?who does not disting- uish between e??a? + pr?s. part, and e??a? + perf. part.?is also the passage Hdt. Ill 89, 3 which, however, may bear another

interpretation: ?p? ?a? ????? ?????t?? . . . ?? ?atest???? ??d??

f???? p???, ???a d??a ????e??. As ?atest???? is not rarely elliptic- ally used to denote the idea of "the existing laws or usages" (Hdt. I 59; Plato Leg. 798 B; Isocr. 7, 56) the sentence may rather be considered an example of 'improper periphrasis'. However, ?? ?. constitutes the 'temporal frame', and the co-existence of combinations containing participles of various degrees of substantivation

may have stimulated speakers and authors to resort to a periphrasis. A parenthesis containing a perfect participle such as IX

102, 3 '????a??? . . . (?dt?? ?a? ?sa? ?? ?pe??? teta??????) s??ep?sp?-

?e??? s??es?p?pt?? ?? t? te???? does not essentially differ from the inserted clauses examined by Bj?rck. Other passages exhibiting a perfect participle with ?? which refers to a state of affairs existing at the moment at which the events narrated came to pass are e.g. Hdt. I 68, 6; II 168, 1. Without questioning the correctness of

Legrand's translation of IV 32 e?e??? ?a? ?? ?a? S???a? . . . ???' ?s??d?? ??? est? pe?? ?pe??????? e??????a, est? d? ?a? ?????? ("c'est chez ?. qu'il est question des H., c'est aussi chez H.") there seems to be room for the observation that an interpretation

?) See P. Chantraine, Histoire du parfait grec (Paris, 1927), 247 f. ; Schwy- zer, Griech. Grammatik, I, 617; 812.

2) Cf. also B. L. Gildersleeve, Syntax of Classical Greek, I (New York, 1900), 122 ff.



A REMARK ON PERIPHRASTIC CONSTRUCTIONS IN GREEK III

of this sentence, or of its 'prototype' as nominal would not be

illegitimate1). A similar remark applies to the 'durative' con-

struction containing those forms of e??a? which were as a rule

not omitted : Soph. Tr. 941 f. ??a??? ?????e?' ?? d???? es???' ??a / . . .

??fa??s????? ????. Periphrasis by means of a perfect participle seems to be equivalent to an adjective in cases such as Hdt. I

199, 5 6sa? ??? ??? e?de?? te ?pa????a? e?s? ?a? ?e???e??, ta?? apa?-

??ss??ta?, dsa? de ????f?? a?t??? e?s? . . . p??s?????s? ; Thuc. II

4, 5? The sentence Hdt. I 103, ? p?? t?? d? ??a??? ?? p??ta ??????

??apef?????a?which was explained by Stein as: pa. ?. ?? ?. ??ap.?

may at least partially go back to an Original' construction p?? t. d. ?.

(predicate) ?? p., ?. ?. (apposition). It is moreover a serious imperfection of both studies, that by

Rosen as well as the book of the Swedish author, that they have

practically disconnected their observations about the periphrastic forms from the history of the perfect. In ancient times the plus-

quamperfect not only was a past tense beside the ancient perfect :

?????e "is glad": ?e???e? "was glad"; t?????e "is dead": ?te????e? "was dead", but it could also indicate that a state or situation had

commenced in the past and continued to exist while other processes came to pass : ? 410 ???' ? ??? ?d? ???? da?e?? ??d?sde ?e???e?, /

??st?? ad t?t' ?f??e Ge?????? ; t 539 0<? d' ??????t? / a????? e? ?e??????, ? d' ?? a????a d?a? ????? 2). As soon as, however, the plusquamperfect became usual to describe pregnantly3), or rather, in a 'fait-ac-

compli Darstellung' 4) an occurrence belonging to the past as if

it had already pre-existed as the result of a former process?cf. Hdt. I 79 ?? d? ?? (Cyrus) ta?ta (a rapid march to Sardes) ed??e, ?a?

?p??ee ?at? t???? ? ???sa? ?a? t?? st?at?? ?? t?? ??d??? a?t?? ???e???

????s?? ??????ee ("he had already come there and was there"); Xen. Hell. VII 2, 9; Anab. V 2, 15 ?ste ??as??? . . . ?????, ?a?

?) For the construction compare also Schwyzer-Debrunner, Griechische Grammatik, II, 119.

2) Compare also contexts such as E 696 t?? d' ???pe ????, ?at? d'

?f?a???? ????t* ????? "was already shed"; e 294; see also ?? Chantraine, Grammaire hom?rique, II, 200; J. Humbert, Syntaxe grecque2 (Paris, 1954) 150; K?hner-Gerth, o.e., I, 152.

3) Schwyzer-Debrunner, o.e., II, 288.

4) W. Havers, Handbuch der erkl?renden Syntax (Heidelberg, 1931), 41 f.; 219 f.



112 A REMARK ON 'PERIPHRASTIC CONSTRUCTIONS IN GREEK

????? e???e, ?a? ????? ??e?e???e?, ?a? ????e? t? ?????? *)?, it began to lose its fitness to fulfil the function studied, as far as the peri-

phrastic forms are concerned, by Bj?rck 2). In his desire to give a

vivid and graphical description of the historical events an author

sometimes tends to pass beyond the limits of the above case of the

perfect participle with the past of e??a?. The type represented by Hdt. IV 137, 3 '?st?a??? de ?????? ta?t?? ?p?de????????? a?t??a

p??te? ?sa? tet?a?????? p??? ta?t?? t?? ??????, p??te??? t?? ???t??-

de? a??e??e??? may be considered to have originated not only in the

trend to replace the 'difficult' 3) forms in -ata? or -at?, but also in

the tendency to substitute, in a vivid narrative and under the

influence of the anticipative activity of phantasy, a resultative

or static verbal category for a narrative historical past 4). It should, moreover, be borne in mind that a periphrastic con-

struction was a necessity in those frequent cases in which the

author wanted to express a modification of the idea expressed by the copula e??a??e.g. ''prove to be, turn out"?, for instance Soph. Tr. 386 ?? ??? ?????? /t??? ??? pa???s?? ??pep??????? ????; 399>

413 ? These combinations have no doubt promoted the spread of

periphrases with e??a?, the more so as ???e?? etc., like e??a?, occurs

also without a participial complement.

Utrecht, van Hogendorpstraat 13.

?) "Die lebendigste, durch das Polysyndeton der kurzen S?tze noch gesteigerte Form der das Resultat vorweggreifenden Erz?hlung; denn das Plqpf. bezeichnet T?tigkeiten oder Zust?nde, welche vor anderen der Vergangenheit angeh?rigen schon vollendet und eingetreten und bis auf den Zeitpunkt dieser ihre Wirkung erstreckt hatten" (Rehdantz-Carnuth, o.e., 71 f.).

2) For particulars concerning the history of the Greek perfect see J. Wackernagel, Studien zum griechischen Perfektum (G?ttingen, 1904); Chan- traine, Histoire du parfait grec (see above).

3) Rosen, o.e., 136, n. g. 4) Cf. Havers, I.e. ; for the Latin periphrasis with habeo see (M. Leumann-)

J. B. Hofmann, Lateinische Grammatik (M?nchen, 1928), 561. E. Schwyzer, Griechische Grammatik, I, does not mention this aspect of the development.



Documents

Gonda Periphrastic Forms