Ancient Greek grammar

From Infogalactic: the planetary knowledge core
Jump to: navigation, search

Lua error in package.lua at line 80: module 'strict' not found. Ancient Greek grammar is morphologically complex and preserves several features of Proto-Indo-European morphology. Nouns, adjectives, pronouns, articles, numerals and especially verbs are all highly inflected. This article primary discusses the morphology of Attic Greek.


<templatestyles src="Module:Hatnote/styles.css"></templatestyles>

The Classical Greek script did not use accents. Accents were devised in the Hellenistic era by scholars who wanted to make it easier for foreigners to learn Greek. The general use of these accents began during the Byzantine Empire. Modern Greek has used only two diacritics since 1982, namely the diaeresis and the acute.

The Ancient Greek script has seven diacritics (two breathings, three accents, a trema (diarhesis), and the Iota subscript).


  • The rough breathing (Greek: δασεῖα, Latin: spiritus asper) (), written over a vowel letter or ρ, marks the sound /h/ at the beginning of a word, preceding the vowel. At the beginning of a word, the vowel υ and the letter ρ always have the rough breathing.
  • The smooth breathing (Greek: ψιλή, Latin: spiritus lenis) (᾿) marks the absence of the /h/ sound.


<templatestyles src="Module:Hatnote/styles.css"></templatestyles>

  • The acute accent (Greek: ὀξεῖα) (΄) is used on long or short vowels or at the third syllable from the end. If there is a long vowel before a long vowel it will take an acute accent. e.g.: κώμη (kṓmē "village"), ἄνθρωπος (ánthrōpos "human").
  • The grave accent (Greek: βαρεῖα) (`) is used on long or short vowels and replaces the acute accent but only on the last syllable. However it is not used when the next word causes an inclination of the accent and also when a punctuation mark follows. e.g.: ὁ καλὸς ποιμὴν ποιμαίνει (ho kalòs poimḕn poimaínei "the good shepherd herds") but ἔλαφός τις (élaphós tis "a deer")—the word τις is behaving as if it is one word with ἔλαφος—and ἐλθέ, Ἰωάννη (elthé, Iōánnē "come, John").
  • The circumflex (Greek: περισπωμένη), displayed as either a tilde (˜) or an inverted breve ( ̑) is used on long vowels. It is on long syllables before short syllables and on nouns in the genitive and the dative case whose final syllable is accented. Also, it is often placed where a contraction has taken place: e.g.: τιμάειν > τιμᾶν (timân "to honor"); κῆπος (kêpos "garden"); nominative αὐγή (augḗ "dawn"), genitive αὐγῆς (augês) and αὐγῶν (augôn), dative αὐγῇ (augêi) and αὐγαῖς (augaîs).




<templatestyles src="Module:Hatnote/styles.css"></templatestyles>

In Ancient Greek, all nouns, including proper nouns, are classified according to grammatical gender as masculine, feminine or neuter and present forms in five distinct morphological cases (nominative, genitive, dative, accusative and vocative). Furthermore, common nouns present distinct forms in the singular, dual and plural number. The set of forms that any particular noun will present for each case and number is determined by the declension that it follows. The form of the declension is additionally determined by the final letter or letters in the stem.

Definite article

Attic Greek has a definite article, but no indefinite article. The definite article agrees with its associated noun in number, gender and case. Proper names usually take a definite article, as do abstract nouns. Adjectives are either placed between the article and noun or after the noun, in which case the article is repeated before the adjective. Dependent genitive noun phrases are positioned in exactly the same way, even though this frequently results in splitting the article and noun by a long dependent phrase. For example, τὸ τοῦ ἀνθρώπου ἔργον tò toû anthrṓpou érgon, literally "the (of the man) deed", or "The deed of the man." In earlier Greek, for instance Homeric Greek, there was no definite article as such, the corresponding forms still having their original use as demonstrative pronouns.

The definite article is declined thus:[1][2]

Masculine Feminine Neuter
Singular Dual Plural Singular Dual [ar 1] Plural Singular Dual Plural
Nominative (ho) τώ (tṓ) οἱ (hoi) (hē) τώ (tṓ) αἱ (hai) τό (tó) τώ (tṓ) τά (tá)
Genitive τοῦ (toû) τοῖν (toîn) τῶν (tôn) τῆς (tês) τοῖν (toîn) τῶν (tôn) τοῦ (toû) τοῖν (toîn) τῶν (tôn)
Dative τῷ (tôi) τοῖν (toîn) τοῖς (toîs) τῇ (têi) τοῖν (toîn) ταῖς (taîs) τῷ (tôi) τοῖν (toîn) τοῖς (toîs)
Accusative τόν (tón) τώ (tṓ) τούς (toús) τήν (tḗn) τώ (tṓ) τάς (tás) τό (tó) τώ (tṓ) τά (tá)
  1. The forms τά (tā́) and ταῖν (taîn) for feminine duals also exist.[2]


<templatestyles src="Module:Hatnote/styles.css"></templatestyles>

Verbs have four moods (indicative, imperative, subjunctive and optative), three voices (active, middle and passive), as well as three persons (first, second and third) and three numbers (singular, dual, and plural). Verbs are conjugated in four main combinations of tense and aspect (present, future, perfect, and aorist), with a full complement of moods for each of these main "tenses", except for the following restrictions:

  • There is no future subjunctive or future imperative.
  • There are separate passive-voice forms (distinct from the middle) only in the future and aorist.

In addition, for each of the four "tenses", there exist, in each voice, an infinitive and participles. There is also an imperfect indicative that can be constructed from the present using a prefix (the "augment") and the secondary endings. A pluperfect and future perfect indicative also exist, but are rather rare. The distinction of the "tenses" in moods other than the indicative is predominantly one of aspect rather than time. The Ancient Greek verbal system preserves nearly all the complexities of Proto-Indo-European (PIE).

A distinction is traditionally made between the so-called athematic verbs, with endings affixed directly to the root (also called mi-verbs) and the thematic class of verbs that present a "thematic" vowel /o/ or /e/ before the ending. All athematic roots end in a vowel except for /es-/ "be". The endings are classified into primary (those used in the present, future, perfect and rare future perfect of the indicative, as well as in the subjunctive) and secondary (used in the aorist, imperfect, and pluperfect of the indicative, as well as in the optative). Ancient Greek also preserves the PIE middle voice; the passive voice that occurs in the future and aorist is an innovation.

Dependence of moods and tenses


The Ancient Greek infinitive is a non-finite verb form with no endings for person, number, tense and mood. It is used mainly to express acts, situations and in general "states of affairs"[3] that are depended on another verb form, usually a finite one. It is a non declinable nominal verb form equivalent to a noun, and expresses the verbal notion abstractly; used as a noun in its main uses, it has many properties of it, as it will be seen below, yet it differs from it in some respects:[4]

(a) When used with no article, and in its major uses (subject/object), it can normally only be equivalent to either a nominative or an accusative case; used with the article, it may be in any case (nominative, genitive, dative and accusative).
(b) It shows morphological formation according to aspect, voice (active, middle, passive) and tense (only the future infinitive).
(c) It retains some verbal syntactic features: it governs the same oblique case (:its object) as the verb to which it belongs, and it may have a subject of its own, in accusative case (See the section accusative plus infinitive below).
(d) It is modified by adverbials and not by adjectives.

Note: an infinitive is also used in independed structures of the following short:

a. Absolutively, with no connexion to any other verb form, in parenthetical expressions.[5]
b. As an adhortative, imperative or prohibitive (negated with μἠ) mood (mostly in poetry, occasionally in proze and mainly in law quotings).[6]
c. Expressing a wish [7] or an exclamation [8] (mostly in poetry).

The infinitive with and without the article

Ancient Greek has both (a) the infinitive with the article (articular infinitive) and (b) the infinitive without the article.

a) The articular infinitive corresponds to a cognate verbal noun. It is preceded by the neuter singular article (τό, τοῦ, τῷ, τό) and has the character and function of both a noun and a verbal form. It can be used in any case (nominative, genitive, dative, accusative) and thus participate in a construction just like any other noun: it can be subject, object (direct or indirect), predicate (rarely), apposition; it may have an adnominal or adverbial use (e.g. to be in a genitive construction as a possessive or objective genitive or it can form a genitive that denotes cause); it may form an exclamation (in poetry); it can also be the complement (:object) of a preposition in any oblique case and denote many adverbial relations; finally, if in the genitive case, it can denote purpose, oftener a negative one.
κακόν ἐστι τὸ παρανομεῖν (the articular infinitive is somehow equivalent to the noun ἡ παρανομία, and in nominative case it functions as the subject of the verb ἐστί)
It is bad to break the law(s) (= Transgression of law is a bad thing)
τοῦτό ἐστι τὸ ἀδικεῖν (as a predicate adjective in nominative), τὸ πλέον τῶν ἄλλων ζητεῖν ἔχειν (nominative apposition to the demonstrative pronoun τοῦτο)
This is injustice: to seek to have more than other people.
μέγα φρονεῖ ἐπὶ τῷ δύνασθαι λέγειν (dative complement of the preposition ἐπί, denoting cause)
He takes pride in being able to speak (= his ability to speak)
τὸ σιγᾶν κρεῖττόν ἐστι τοῦ λαλεῖν (genitive of comparison depended on the comparative degree adjective κρεῖττον)
Keeping silence is better than speaking
ἐτειχίσθη Ἀταλάντη... τοῦ μὴ ληστὰς κακουργεῖν τὴν Εὔβοιαν (genitive of purpose)
Αtalante was fortified to prevent pirates from ravaging Euboea.
b) The infinitive without the article is of two sorts and has two discrete uses: the dynamic infinitive and the declarative infinitive [9][10] (traditionally they are also said to be used not in indirect discourse and in indirect discourse respectively.[11])
Dynamic vs. declarative infinitive
Dynamic infinitive

A so-called dynamic infinitive may be governed by verbs of will or desire to do something (ἐθέλω or βούλομαι "to be willing, wish to" αἱροῦμαι "choose, prefer to", μέλλω "to be about to, or: delay to", κελεύω "urge, command to", ἐπιτάσσω "order to", ψηφίζομαι "vote to", ἐῶ "allow to", δέομαι "beg to" etc.), verbs of will or desire not to do anything (δέδοικα/δέδια "fear to", φοβοῦμαι "be afraid to", ἀπέχομαι "abstain from doing", αἰσχύνομαι "be ashamed to", ἀπαγορεύω "forbid to", κωλύω "hinder, prevent" etc.) and verbs or verbal expressions denoting ability, fitness, necessity, capacity, etc. (δύναμαι, ἔχω "be able to", ἐπίσταμαι, γιγνώσκω "know how to", μανθάνω "learn to", δυνατὸς εἰμί, ἱκανὸς εἰμί "I am able to", δίκαιον ἐστί "it is fair/right to", ἀνάγκη ἐστί "it is necessary to", ὥρα ἐστί "it is time to" etc.). It can also be found after adjectives (and sometimes derived adverbs) of kindred meaning (δεινός "skillful", δυνατός "able", οἷός τε "able", ἱκανός "sufficient, capable" etc.). It stands as the object (direct or indirect) of such verbs or verbal expressions, or it serves as the subject if the verb/the verbal expression is used impersonally; it also defines the meaning of an adjective almost as an accusative of respect. An infinitive of this kind denotes only aspect or stage of action, not actual tense,[12] and can be in any tense stem (mostly in the present and aorist (see also here), the perfect being rare enough) except the future one (only the verb μέλλω "I am about to" may exceptionally take a dynamic future infinitive).[13]

The difference between the present and the aorist infinitive is aspect or stage of action, not the tense —despite their tense stem, such infinitives always have a future reference, because of the volitive meaning of their governing verb. More specifically, an infinitive in the present verb stem lays stress on "the process or course of the state of affairs", and in many cases has "an immediative" semantic force, while an infinitive in the aorist verb stem lays stress "on the completion of the state of affairs, expressing a well-defined or well-delineated state of affairs".[14]

Present dynamic infinitive (continuing stage of action):
βούλομαί σε εἰς τὰς Ἀθήνας ἰέναι.
I want (for) you to go (=every time, or=to start/keep going to Athens etc.).
ἀνάγκη ἐστὶ μάχεσθαι.
It is necessary to fight (= to start/keep fighting).
Aorist dynamic infinitive (completed stage of action):
βούλομαί σε εἰς τὰς Ἀθήνας ἐλθεῖν.
I want (for) you to go to Athens. (just once, a simple and sole occurrence of going)
αἰσχύνομαι ὑμῖν εἰπεῖν τἀληθῆ.
I'm ashamed to tell you the truth. (just once, a simple occurrence of telling, as in the previous example)
Declarative infinitive

A so-called declarative infinitive [15] (see also declarative sentence) is mostly used in connexion with verbs (or verbal expressions) of saying, thinking and (sometimes) perceiving such as λέγω, φημί, ἀποκρίνομαι, ὑπισχνοῦμαι, ὁμολογῶ, ἀκούω, ὁρῶ etc. and it is usually used in oratio obliqua (in indirect speech or discourse). The latter means that it represents a corresponding finite verb form of the oratio recta (of the direct speech or discourse),[16] thus a declarative infinitive denotes both tense and aspect or stage of action. But the present infinitive represents either a present indicative or an imperfect one, and a perfect infinitive either a perfect indicative or a pluperfect one. A declarative infinitive with the particle ἂν can also be a representative of a potential indicative or potential optative of the corresponding tense.[17]

ᾤετο ἀποθανεῖσθαι. Future infinitive = future indicative.
He was thinking that he would die. (direct form: ἀποθανοῦμαι "I will die". )
ᾤοντο ὑμᾶς ἀποθανεῖν. Aorist infinitive = aorist indicative.
They were thinking that you died. (direct form: ἀπέθανον "they died".)
ἔφην σε ἀφικνεῖσθαι. Present infinitive = either present or imperfect indicative, according to context.
I said that you were arriving (at the very moment), or: I said that you had been arriving (some time ago) (direct forms: ἀφικνεῖται "He is arriving" and ἀφικνεῖτο "He was arriving", respectively).
...οὓς (anaphoric to τοὺς καλοὺς ἐκείνους καὶ μακροὺς λόγους) ἐν τοῖς μυρίοις ἐν Μεγάλῃ πόλει πρὸς Ἱερώνυμον τὸν ὑπὲρ Φιλίππου λέγοντα ὑπὲρ ὑμῶν ἔφη δεδημηγορηκέναι.
... which (anaphoric to "the fine these long orations") he said he had delivered as your spokesman before the Ten Thousand at Megalopolis in reply to Philip's champion Hieronymus. (the perfect infinitive δεδημηγορηκέναι can represent either a perfect indicative δεδημηγόρηκα "I have delivered orations" or a pluperfect one ἐδεδημηγορήκειν "I had delivered orations", the interpretation being left exclusivelly on contextual parameters)
ἔφασαν ἂν ἀποθανεῖν, εἰ τὰ ὅπλα παραδοῖεν. Potential infinitive = potential optative.
They said that they might die, if they surrender their arms (direct form: ἂν ἀποθάνοιμεν, εἰ... "we may die if...".).
ἔφασαν ἂν ἀποθανεῖν, εἰ τὰ ὅπλα παρέδοσαν. Potential infinitive = potential/unreal (conditional) indicative.
They said that they would have been dead, if they had surrendered their arms. (direct form: ἂν ἀπεθάνομεν, εἰ... "we would have died if...".)

Verbs that usually have a future reference, such as ὄμνυμι "swear", ὑπισχνοῦμαι "promise", ἐλπίζω "hope", ἀπειλέω "threaten", προσδοκάω "expect" etc., either take the declarative infinitive (mostly the future, but less often also the present, aorist or perfect infinitive, even the infinitive with ἀν representing a potential optative or indicative), and in this case indirect discourse is employed, or they are followed by the dynamic aorist (less often the present) infinitive, and they are constructed just like any verb of will, desire etc.[18][19] The same constructional alternation is available in English (declarative content clause -a that clause- or to-infinitive), as shown below.

ὄμνυμι τὰ χρήματα ἀποδώσειν.
I swear that I will give back the money.
ὄμνυμι τὰ χρήματα ἀποδοῦναι.
I swear to give the money back.

But in the last example another reading is also possible, considering ἀποδοῦναι to be an aorist declarative infinitive : "I swear that I gave the money back."

The infinitive in subordinate clauses introduced by conjunctions

The ("dynamic") infinitive is used instead of the indicative mood, with substantial difference in meaning, in certain subordinate clauses introduced by specific conjunctions: ὥστε (ὡς) "so as to, so that",[20] πρίν (πρόσθεν... ἤ) "before" or "until"[21] and relative adjectives introducing relative clauses of result, such as ὅσος "so much as enough to", οἷος "of such a short as to",[22] ὃς or ὅστις "(so...) that he could", in clauses introduced by the prepositional prhases ἐφ' ᾧ or ἐφ' ᾧτε or with ὥστε "with the proviso that".[23] Note: a "declarative" infinitive is sometimes the mood of subordinated clauses in indirect speech, instead of a corresponding indicative (either a realis or conditional irrealis one) or optative mood, in modal assimilation to the main infinitive used to represent the independed clause of the direct speech; so after relative, temporal or conditional conjunctions such as: ὃς "who" or ὅστις "whoever", ἐπεὶ or ἐπειδή "since, when", ὅτε "when", εἰ "if" etc.[24]

Subject of the infinitive

In general, Greek is a pro drop language or a null-subject language: it does not have to express the (always in nominative case) subject of a finite verb form (either pronoun or noun), unless it is communicatively or syntactically important (e.g. when emphasis and/or contrast is intended etc.). Concerning infinitives, the following can be said as an introduction to the infinitival syntax (case rules for the infinitival subject):

When the infinitive has a subject of its own (that is, when the subject of the infinitive is not co-referential either with the subject or the object of the governing verb form), then this word stands in the accusative case. When the subject of the infinitive is the same as the subject of the main verb (when the subjects are co-referential), then it is usually neither expressed nor repeated within the infinitival clause. This omission is noticed also when the subject of the infinitive, although not the subject of the governing verb, is constructed in a higher syntactic level, e.g. as (any) object of the governing verb in any oblique case, or as a dative of interest with an impersonal verb or verbal expression. But even in this case an accusative may be present or—more usually—understood by a predicate adjective, participle etc. in the accusative.[25] These three main constructions available are desctribed in some detail in the sections below.
Accusative plus infinitive

The construction where an accusative argument belongs exclusively to the infinitive and functions as its subject is called accusative plus infinitive (see also the Latin construction accusativus cum infinitivo (ACI). Yet, in Latin this construction is the rule even in cases where verb and infinitive have co-referential subjects). This construction can be used as an indirect speech mechanism, in many occasions interchangeable with a complementary declarative clause introduced by "ὅτι/"ὡς" (or a supplementary participle). But with some verbs (normally with verbs of thinking, as νομίζω, οἴομαι, ἡγέομαι, δοκέω etc., with the verb φημί "say, affirm, assert", with verbs denoting hope, oath or promise, such as ἐλπίζω "hope", ὄμνυμι "swear", ὑπισχνοῦμαι "promise", etc.) the infinitival construction is the rule in classical Greek. But it can be also in use with any infinitival use, no matter whether indirect speech is involved or not. In the following examples the infinitival clause is put in square brackets [] (the first translation given is a literal rendering, the second being a real English one):

λέγουσίν τινες [Σωκράτη σοφὸν εἶναι].
say some [Socrates ACC wise ACC be INF] (Subject and predicate adjective are in accusative case)
Some people say that Socrates is (or: was) wise.
νομίζουσιν [Σωκράτη σοφὸν εἶναι].
pro pl 3rd think [Socrates ACC wise ACC be INF] (as stated before)
They consider Socrates to be (or: to have been) wise.
Oratio Recta/Direct speech for both the above examples: "Σωκράτης σοφός ἐστιν (or ἦν). Socrates NOM is (or: was) wise NOM." (Subject and predicate adjective of the finite verb ἐστί in nominative case)

This construction is also the rule when the main verb form is impersonal/an impersonal expression and the infinitive functions as its subject (no indirect speech).

[ὑμᾶς ACC] χρὴ [τὴν αὐτὴν γνώμην ἔχοντας ACC τὴν ψῆφον φέρειν].
[You ACC] is necessary [the same opinion having ACC the vote carry INF] (Subject of the infinitive is in accusative; the participle ἔχοντας shows also concord with the acccusative pronoun ὑμᾶς)
It is necessary for you to vote having the same opinion in mind.

Note: The accusative plus infinitive is sometimes also used, normally for the sake of emphasis or contrast, even in coreferential subject structures; then the accusative of a personal (in the first and second persons, less often in the third) or a reflexive pronoun (in the third person) may be used.[26]

Nominative plus infinitive

When the subject of the infinitive is identical (coreferential) with the subject of the governing verb, then normally it is omitted and understood in the nominative case. The phenomenon is traditionally understood to be some kind of case attraction [27] (for a modern perspective and relevant modern terminology see also big PRO and little pro). In the following examples infinitival clauses are bracketed [].

proi νομίζετε [PROi τοὺς πολεμίους ἂν νικῆσαιINF]. (Potential infinitive)
proi think-you [PROi the enemies can-beatINF] (literal translation)
You think (that) you can/could beat your enemy. (idiomatic translation)
Oratio Recta/Direct speech: "Τοὺς πολεμίους ἂν νικήσαιμεν. We could beat our enemy." (Potential optative)
[Πέρσηςi NOM] proi ἔφη [PROi εἶναιINF].
[Persiani NOM] proi said-he [PROi beINF] (literal translation)
Hei said (that) hei was (a) Persiani. (idiomatic translation)

Πέρσης is a predicate noun in the nominative, showing case agreement with an understood and omitted pronoun (coreferential proi and PROi).

Oratio Recta/Direct speech: "Πέρσης εἰμί. I am (a) Persian".

Note 1: Interestingly, this construction is preferred even when an impersonal infinitival construction intervenes. As it was stated in the previous section about the use of (finite) impersonal verbs that require an infinitive as their first argument (:subject), normally an accusative plus infinitive construction would be expected.

pro φημὶ δὴ [χρῆναιINF [πειρᾶσθαιINF [σῴζεινINF τοὺς ἄνδρας καὶ μὴ μνησικακεῖνINF, ἐνθυμουμένουςACC ὅτι πολλὰ καὶ ὑμεῖς ὑπὸ τῶν ἐπιβουλευσάντων ἐξηπάτησθε, [[[ὧν οὐδενὸς αὐτοὶj NOM δοῦναιINF δίκην] δίκαιον εἶναιINF] proj ἂν φήσαιτε]]]].
pro say-I [is-necessaryINF [tryINF [saveINF the men and not remember-past-injuriesINF, rememberingACC that..., [[of-which of-nothing selvesj sufferINF penalty] right beINF] proj could-admit-you]]]].
I say that it is your duty to try to save these men and to forget past injuries, remembering that you too have in many cases been led by plotters in errors, for none of which you would admit that it is right for you to pay yourselves the penalty.

After the governing verb φημί "I say", the infinitives πειρᾶσθαι, σῴζειν, μὴ μνησικακεῖν share an omitted and understood accusative ὑμᾶς as their subject (here the omission is rather due to the existence of the pronoun ὑμεῖς in the following ὅτι clause); that is why the participle ἐνθυμουμένους stands in the accusative. Yet, in the relative clause with ἂν φήσαιτε "you would admit" as governing verb, the subject of the infinitive δοῦναι is understood in the nominative, as the also in nominative definite pronoun αὐτοὶ shows, which is here used reflexively and in coreference with the subject ὑμεῖς "you" (implied in the second person plural ending of the verb φήσαι-τε), even though an infinitival impersonal expression intervenes (δίκαιον εἶναι "it is fair") and accusative could be used, as in the not coreferential main clause structure with the verb φημί. So, without the governing verb ἂν φήσαιτε (i.e. in direct speech) the construction would be:

[Τούτων οὐδενὸς] δίκαιόν ἐστιν [ἡμᾶςACC ('αὐτοὺςACC) δοῦναιINF δίκην].
[Of-these of-nothing] right is [usACC (selvesACC) sufferINF penalty]
It is not fair for us to pay (ourselves) the penalty for any of these (errors).

Yet, with the governing verb ἂν φήσαιτε "you could admit" introducing indirect speech mechanism (for convenience placed now at the beginning of the sentence) the construction has become (the infinitival clause is bracketed [...]):

ἂν φήσαιτε [τούτων οὐδενὸς] δίκαιόν εἶναι INF [αὐτοὶ ΝΟΜ δοῦναι INF δίκην].
proi could-admit-you [of-these of-nothing] right beINF [PROi αὐτοὶi NOM δοῦναιINF δίκην].
You would admit that for any of these (errors) it is not right for you to pay yourselves the penalty.

Note 2: with many impersonal verbs and impersonal expressions, or with passivized verbs that are (or they should be) impersonally used, it is common the raising-to-subject construction. This phenomenon was very perceptively pointed out already by Raphael Kühner,[28] and from then and on is emphasized by many grammarians.[29] Such verbs or verbal expressions are: λέγεται "it is said that", ἀγγέλλεται "it is announded that", ὁμολογεῖται "it is agreed/granted that", δοκεῖ "it seems that", συμβαίνει "it happens that", δίκαιόν ἐστιν "it is fair to", ἄξιόν ἐστιν "it is worthy to" etc.

Some examples are the following:

δοκεῖς δέ μοι [καὶ αὐτὸςNOM τοῦτο γιγνώσκειν].
seem-you proi [too PROi selfNOM i this recognizeINF] (literal translation)
You seem to me to recoginze this yourself too. (idiomatic translation)

Here 2nd person (= σὺ "you") subject in nominative is pro-dropped; the definite, reflexively used, pronoun αὐτὸς shows nominative morphology, agreeing with the omitted subject.

Λέγεται τοίνυν [ποτ' ἐν τῇ πόλει κατὰ τὴν παλαιὰν ἐκείνην εὐδαιμονίαν] ἈλκιβιάδηςNOM [γενέσθαιINF].
is-said moreover [at-some-time in the city in the old that prosperity] Alcibiadesi NOM [PROi liveINF] (literal translation)
Moreover, Alkibiades is said to have lived in the city (of Athens) in the good old days of its prosperity. (Or: It is said that Alcibiades had lived...) (idiomatic translation)

Compare the above with the following, constructed in the "accusative plus infinitive" way:

λέγεται γὰρ [ἈλκιβιάδηνACC, πρὶν εἴκοσιν ἐτῶν εἶναι, Περικλεῖ... τοιάδε διαλεχθῆναιINF περὶ νόμων].
Is-said for [Alcibiades ACC... have-a-talkINF about laws with Pericles] (literal translation)
For it is said that Alciviades, before becoming twenty years old, had a talk about laws with Pericles. (idiomatic translation)

In both of the above sentences a good translation could be "A history is told about Alcibiades, that he..." or "History tells (us) that Alcibiades..." [30]

Yet, in the following sentences raising-to-subject is inapplicable, since another infinitive used impersonally (δεῖν, δίκαιον εἶναι) intervenes between the impersonal verb δοκεῖ μοι and the personally used infinitive:

δοκεῖ μοι [δεῖνINF [καὶ τοὺς μὴ γεγονόταςACC ἤδη δεδιέναιINF περὶ τῶν μελλόντων ἔσεσθαι]]. (The accusative is the subject of δεδιέναι)
Seems to-me [necessary-beINF [even the-not-bornACC yet anxious-feelINF about the future events]]. (literal translation)
It seems to me that it is necessary even for our unborn yet fellow citizens to be in anxious awareness of the future events. (idiomatic translation)
πολὺ δὲ μάλιστα δοκεῖ ἡμῖν [δίκαιον εἶναιINF, [εἴ τις ἡμῶν αὐτῶν λυμαίνεται ταύτῃ τῇ καταστάσει, δίκην αὐτὸνACC διδόναιINF]].
much but above-all seems to-us [right beINF, [if anyone of-us selves is-harming this the situation, punishment he-selfACC sufferINF]] (literal translation)
But in particular we consider it to be right that, if anyone of our own number is harming this order of things, he should be punished. (idiomatic translation)

Here a raising-to-subject construction would be ungrammatical (for example, in the first of the above sentences):

*δοκοῦσι μοι καὶ *οἱ μὴ γεγονότεςi NOM... [δεῖνINF [δεδιέναιINF περὶ τῶν...]].
  • seem-they to-me even-too the-not-borni NOM [necessaty-beINF] [[anxious-feelINF about the...]]. (literal translation)
  • Even our unborn citizens seem to me that it is necessary to be in anxious awareness of the...

The phrase after δοκεῖ μοι, except the transformation of the embedded clause from indicative to the infinitive mood, is structured exactly as if this verb was not present (e.g. δεῖ καὶ τοὺς μὴ γεγονόταςACC ἤδη δεδιέναιINF περὶ τῶν μελλόντων ἔσεσθαι). It seems that in cases like these at hand, the whole expression composed by both the impersonal verb and the impersonally used infinitive prevents raising-to-subject.

Note 3: With some verbs the raising-to-subject construction, if only not an impersonal infinitive follows, is the rule (like δοκεῖ), while with some others (like λέγεται) it is just preferred.

Note 4: The raising-to-subject construction is also applicable to the supplementary participial clauses, see below.

Subject omitted and understood in an oblique case (genitive, dative or accusative)

When the infinitival subject is coreferent with a word constructed with the governing verb in a higher syntactic level, in other words, when the subject of the infinitive is itself (a second) argument of the governing verb, then it is normally omitted and understood either in the oblique case in which the second argument is put, or in the accusative as in any other accusative plus infinitive construction.

νῦν σοι DAT ἔξεστιν [ἀνδρὶ DAT γενέσθαι].
now for-youDAT is-possible [man DAT to-become INF]. literal translation
Now it is possible for you [to become a man]. idiomatic translation
(Predicate noun ἀνδρί "a man" in case agreement with the dative σοι "for you".)
ἐδέοντο τοῦ Κύρου GEN [ὡς προθυμοτάτου GEN πρὸς τὸν πόλεμον γενέσθαι].
were-begging-they Cyrus GEN [as-willing-as-possible GEN for war to-be INF]. literal translation
They were begging Cyrus [to be as willing as possible for war]. idiomatic translation
(Predicate adjective προθυμοτάτου "willing" shows case agreement with the genitive object τοῦ Κύρου "Cyrus".)

In both of the above examples the case of the subject of the infinitive is governed by the case requirements of the main verb and the infinitive is appended as a third argument[31] (Concerning the second example, in modern linguistic terms we have to do with an object control construction). Traditionally this construction is sometimes called (in Latin terminology) dativus cum infinitivo or genitivus cum infinitivo (dative with the infinitive or genitive with the infinitive respectively) and is considered to be a case attraction,[32][33] the dative or genitive being used instead of a predicate in the accusative: ἄνδρα, ὡς προθυμότατον; see also below.

On the other hand, as it is shown by predicate adjectives or participial constituents of the infinitive clause, in the following examples an accusative is understood and must be supplied by context as the subject of the infinitive.

Λακεδαιμονίοις DAT ἔξεστιν [ὑμῖν φίλους ACC γενέσθαι].
For-Lacedaemonians DAT is-possible [to-you friends ACC] to-become INF]. literal translation
It is possible for Lacedaemonians [to become friends to you]. idiomatic translation
(Predicate adjective φίλους "friends" is in the accusative, in case agreement with an understood and omitted accusative subject Λακεδαιμονίους "Lacedaemonians".)
δεόμεθ᾽ οὖν ὑμῶν GEN [ἀκροάσασθαι τῶν λεγομένων, ἐνθυμηθέντας ACC ὅτι...]
are-begging-we of-you GEN [to-listen-carefully <subINF of-what-is-being-said, keeping-in-mind ACC that...] literal translation
We are begging you [to listen carefully of what is said, keeping in mind that...] idiomatic translation
(Participle ἐνθυμηθέντας "keeping in mind" in the accusative, agreeing in case with an omitted/understood accusative ὑμᾶς "you".)

This construction is obligatory when the infinitive is governed by a participle in any oblique case, more usually an attributive one (and in the nominative also).[34] So an embedded participial clause like "φάσκοντες εἶναι σοφοί" "claiming that they are wise" or "οἱ φάσκοντες εἶναι σοφοί" "Those who claim that they are wise" is declined this way -in any of the following word ordering, but in slightly different each time meaning/topicalization:

(οἱ) φάσκοντες σοφοὶ NOM εἶναι
(τοὺς) σοφοὺς φάσκοντας ACC εἶναι
(τῶν) φασκόντων GEN εἶναι σοφῶν GEN
(τοῖς) σοφοῖς DAT εἶναι φάσκουσι DAT

In the above phrasal structuring the predicate adjective "σοφοὶ" "wise" is always put in the case of its governing participle "φάσκοντες" "claiming".


The participle is a non-finite nominal verb form (it consists of a non-finite clause else called participle clause or participial clause) declined for gender, number and case (thus, it is a verbal adjective) and has many functions in Ancient Greek. It can be active, middle or passive and can be formatted in present, future, aorist and perfect tense stems. These tense stems denote only stage of action/aspect, and normally not absolute time but only time relatively present, past of future with reference to another verb form, either finite or not-finite.[35][36] In general, as it shows no personal endings, its main use is to express an action, situation etc. that accompanies the action, situation etc. expressed by this other verb form (which means that it is a constituent of it or it modifies any other constituent, thus the one or the other way it depends on it).[37]

Uses of the participle

Three main syntactic uses of the participle can be distinguished: (a) the participle as a modifier of a noun (attributive participle) (b) the participle used as an obligatory constituent (:argument) of a verb (supplementary participle), (c) the participle as an adverbial satellite of a verbal predicate (circumstantial or adverbial participle.[38]

The attributive participle

The attributive participle is often, though not always,[39] used with the article (which can be either generic[40] or particular[41][42]), it functions as a common adjective, it can be in every tense stem, and it is on a par with -and thus often translated as- a relative clause.[43] It shows agreement with a noun, present of implied in a sentence, and can be assigned any syntactic role an adjective can hold.

οὗτός ἐστι ὁ κλέψας τὸν χιτῶνα. (subject)
This is who stole the chiton.
ἔχων ὄνον τις ἐπώλησεν αὐτόν. (an attributive modifier of the subject τις "someone")
Someone who had a donkey sold it.

Like any adjective it can be used substantively (or be nominalized), by omission of an understood noun (easily recoverable from the context):

οἴκαδε βουλόμενος ἀπιέναι (sc. ἀνὴρ man or στρατιώτης soldier)…
Whoever (= any man or soldier who) wants to go home…

Many participles of this sort are equivalent to -and thus translated as- nouns, e.g.:

ὁ φεύγων = the person put on exile, or the person accused, i.e. the exile, or the defendant
τὸ μέλλον = the thing that is going to happen, i.e. the future
ὁ κλέπτων = the person that steals, i.e. the thief

An adverbial notion may be inherent in an attributive participle; usual notions are those of purpose or consequence (in future tense) and condition (in all tense stems but the future, with negation μή).

Νόμον δημοσίᾳ τὸν ταῦτα κωλύσοντα τέθεινται τουτονί.
They have publicly enacted this specific law, which is to prevent (: in order this specific law to prevent) these things.
ὁ μὴ δαρεὶς οὐ παιδεύεται.
Whoever is not unflogged (:if anyone is unflogged) can’t be trained.
The supplementary participle

The supplementary participle is always without the article (predicative position [44] of the article is employed) and can be in any tense stem. This participle has two major uses: (1) in indirect discourse, and (2) not in indirect discourse.

The supplementary participle in indirect discourse

When in this use, the participle corresponds to a particular tense and mood of a declarative sentence: simple indicative of any tense, and, if accompanied by the particle ἄν, potential optative or potential indicative. Verbs taking such a participial clause as an object complement are:

i) those meaning sense, knowledge, learning, memory and the contrary verbs such as: ὁράω "see that", ἀκούω, "hear that" αἰσθάνομαι "perceive, understand that", γιγνώσκω "come to know, perceive, realize that", ἀ-γνοέω (= οὐ γιγνώσκω) "not to perceive or recognize that",(κατα)λαμβάνω "find (on arrival) that", εὑρίσκω "find that", αἱρέω "catch, detect someone doing something" (passive: ἁλίσκομαι), μανθάνω "learn, know that", οἶδα "know that", σύνοιδα "know (as a witness), or be conscious that", ἐπίσταμαι "know that", τίθημι "consider that" μέμνημαι "remember that", ἐπιλανθάνομαι "forget that" etc.

τίθημί σε ὁμολογοῦντα, ἐπειδὴ οὐκ ἀποκρίνῃ.
I consider you to agree (:I consider that you agree) with me , since you don’t give me any reply.
καὶ πρόσθεν μὲν οἶδα αὐτοὺς φοβουμένους χρυσίον ἔχοντας φαίνεσθαι.
And I know too that in the past they were afraid to be seen in possession of gold.

In many of the above cases this participle is interchangeable with a complementary declarative clause introduced by ὅτι or ὡς (a sort of an argument clause or else called a content "that" clause). Compare the last example with the following.

τίς γὰρ οὐκ οἶδεν ὅτι καὶ τὴν χώραν ἡμῶν κατανενέμηνται καὶ τὴν πόλιν κατεσκάφασιν;
For who does not know that they (:the Thebans) have portioned out our land for grazing and have destroyed our city to the ground?

ii) verbs of presentation, i.e. verbs meaning "announce, show, prove etc." such as: ἀγγέλλω "report that", (ἀπο-, ἐπι-) δείκνυμι "show, explain, point out that", ἐλέγχω "bring convicting proves that" etc.

The following case constructions are available:

a. The supplementary participle’s subject and the governing verb's subject are coreferential, the participle being put in the nominative case, agreeing with it (we are dealing with a so-called nominative plus participle construction; see also nominative plus infinitive):

ὁρῶσιν [ἀδύνατοι ὄντες περιγενέσθαι]
They see [that they are in no position to get the upper hand].
Direct form: ἀδύνατοί ἐσμεν περιγενέσθαι. We are in no position to get the upper hand.

b. The supplementary participle has a subject of its own (there is no coreference) and both the participle and its subject are put in the accusative case, just like an accusative plus infinitive construction. This is the case where the argument of the verb is not the noun, even though it seems to be an accusative object, but the verbal notion expressed by the participle itself:[45]

αἰσθάνομαί [τινας παραβαίνοντας τοὺς νόμους].
I understand [that some people break the laws. Independent/direct speech form: "παραβαίνουσί τινες τοὺς νόμους". Some people break the law.
ἀπέδειξε [Λύσανδρον κτείναντα Φιλοκλέα].
He proved [that Lysander killed Philocles]. Independent form: "Λύσανδρος ἔκτεινε Φιλοκλέα". Lysander killed Philocles.

In each of the above sentences, if the participle is taken away, then the remaining construction is ungrammatical, considering that each governing verb retains its initial meaning. This proves that the noun (or the noun phrase) is an argument of the participle only, and by no means of the verb.[46]

αἰσθάνομαί* τινας. ‘’I understand* some people’’.
ἀπέδειξε* Λύσανδρον. ‘’He proved* Lysander’’.

It must have been made clear from the above examples that those verbs are bivalent ones (or monotransitive) and not ditransitive, i.e. they have no other semantic argument except their subject and the supplementary participial clause itself.

Compare the English: He proved [Lysander [to have killed Philocles]] vs. He proved [that Lysander killed Philocles]; in modern linguistic terms, in the first phrase structure subject-to-object raising is considered to be employed. It can also be argued that the same stands for the use of this kind of participles in Greek.[47] On the contrary, with verbs like ὁράω "see", ἀκούω "hear", εὑρίσκω "find" etc. there is also possible a construction that has nothing to do with indirect speech, but is a mere description of a sensory input (direct perception):

εἶδον αὐτοὺς [πελάζοντας].
They saw them [approaching]. (Not: They saw [that they were approaching].)
ηὗρε τοὺς ἄνδρας [διεφθαρμένους].
He found the men [already put to death]. (Not: He found out [that the men had already been put to death].)
ἤκουσα αὐτοῦ [καὶ περὶ φίλων διαλεγομένου].
I heard him [discussing about friendship too]. (Not: I heard [that he was discussing about friendship too].)

In the above sentences αὐτούς, τοὺς ἄνδρας and αὐτοῦ are second arguments of the verbs εἶδον, ἤκουσε and ηὗρε respectively, while the participles are added as their third arguments (in modern linguistic terms called (verbal) secondary predicates or small clauses, although almost every supplementary participial clause -see also below- could be classified as such).

Especially with verbs taking either the genitive plus participle or the accusative plus participle construction, the semantic difference between these two options is important; with the "accusative plus participle" an indirect speech mechanism is invoked, while with the "genitive plus participle" only the physical perception of an action is brought into play. But with verbs that (almost) always take exclusively the accusative the distinction must be made only on the ground of some semantic properties of the noun in accusative (e.g. [+visible], [+ findable] etc.), the argument structure of the main verb or/and even by contextual parameters exclusively. Compare the last example with the following, where the noun and the participle are put not into the genitive plus participle but into the accusative plus participle construction (thus, here indirect speech is employed):

ἤκουσε [Κῦρον ἐν Κυλικίᾳ ὄντα].
He heard [that Cyrus was in Cilicia].

NOTE 1: Sometimes, mainly for the sake of contrast of emphasis, in a coreferent structure where we should expect a participle in nominative case, there is an accusative reflexive pronoun standing as subject of the participle, and thus the participle is also put in the accusative.

πρὸς δὲ τούτοις καὶ ἐμαυτὸν (acc.) ἐπέδειξα πάντων μετειληφόθ (acc.) ὅσων προσήκει τοὺς ἐλευθέρους.
And besides these things, I exhibited that I myself have shared in all the privileges befitting free men.

Compare the above with the following:

σωφροσύνης πίστιν περὶ ὑμῶν ἔχοντες (nom.) ἔδειξαν.
They exhibited that they have confidence in you.

The use of the reflexive pronoun is the rule when an accusative plus participle construction or a participle of an impersonal verb precedes and is coordinated with a participle, whose subject is coreferential with the subject of the verb.

πειράσομαι δεῖξαι καὶ μετὸν τῆς πόλεως ἡμῖν καὶ πεπονθότ' ἐμαυτὸν οὐχὶ προσήκοντα.
I shall try to prove that we have legal rights to citizenship, and that I have suffered things not worthy of me.

Here a supplementary participle in an impersonal construction precedes ("μετὸν τῆς πόλεως ἡμῖν" < "μέτεστι τῆς πόλεως ἡμῖν"). In another case, the partiiple could have been structured in the nomimative case:

πειράσομαι δεῖξαι πεπονθὼς οὐχὶ προσήκοντα.

But when the participle with the coreferential subject precedes, then the nominative plus participle construction normally occurs with the first participial clause, while the following stands in the accusative plus participle:

ἑώρων [τοῖς τε ἐπιχειρήμασιν οὐ κατορθοῦντες] [καὶ τοὺς στρατιώτας ἀχθομένους τῇ μονῇ.
They were seeing that [they were not having success in their designs, and that the soldiers also were weary of staying].

Here neither emphasis nor contrast is laid upon, and the first construction is a nominative plus pariticiple while the second an accusative plus participle one.

NOTE 2:Uppon the verbs σύνοιδα and συγγιγνώσκω normally a dative is depended, with which the participle shows agreement. Instead of the dative, occasionally an accusative plus participle construction can also be found.

καὶ τί μοι σύνοισθα, ὦ Σώκρατες, τοιοῦτον εἰργασμένῳ; (μοὶ εἰργασμένῳ is in the dative)
What knowledge, Socrates, can you possibly have of my being guilty of such a thing as that?
νῦν δ᾽ ἐγὼ σὲ σύνοιδα ἐπὶ μὲν κωμῳδῶν θέαν καὶ πάνυ πρῲ ἀνιστάμενον καὶ πάνυ μακρὰν ὁδὸν βαδίζοντα καὶ ἐμὲ ἀναπείθοντα προθύμως συνθεᾶσθαι.(μὲ ἀνιστάμενον, βαδίζοντα, ἀναπείθοντα is in the accusative; there is some manuscript authority for σοὶ (dat.)... ἀνιστάμενον etc. (acc.), yet many scholars emend)
At present I observe that when you are about to see a comedy, you get up very early and walk a very long way and press me eagerly to attend the play with you.

But when the dative is coreferential with the subject, then the reflexive pronoun is used as a second argument, and in this case the participle, used more likely as a third argument, may optionally show agreement with either the nominative subject (the first argument) or the reflexive dative (the second argument).

σύνοιδα ἐμαυτῷ (dat.) πάντα ἐψευσμένος (nom.) αὐτόν. (agreement with the subject in nominative)
I am self-conscious that I have proved utterly false to him.
ἐγὼ γὰρ δὴ οὔτε μέγα οὔτε σμικρὸν σύνοιδα ἐμαυτῷ' (dat.) σοφὸς ὤν (nom.). (as stated above)
For I am self-conscious that I am not wise either much or little.
ἐμαυτῷ (dat.) γὰρ συνῄδη οὐδὲν ἐπισταμένῳ (dat.) ὡς ἔπος εἰπεῖν, τούτους δέ γ᾽ ᾔδη ὅτι εὑρήσοιμι πολλὰ καὶ καλὰ ἐπισταμένους. (agreement with the dative)
For I was self-conscious that I knew practically nothing, but I knew I should find that they knew many fine things.

The same somehow principle stands for the verb ἔοικα (rarely with the periphrasis ὅμοιός εἰμί) "look like, be the same as", the difference being that with these verbs the dative is not present but implied (perhaps an indefinite pronoun form, like τινί or τισίν "someone").

ἐοίκατε τυραννίσι μᾶλλον ἢ πολιτείαις ἡδόμενοι (nom.). (agreement with the implied nominative subject ὑμεῖς)
You seem to take pleasure in despotisms rather than in free governments.
ἔοικας ὀκνοῦντι λέγειν.
You act like one who shrinks from telling his thought.
συνανέστη ὁ Σεύθης οὐδέν τι μεθύοντι (dat.) ἐοικώς.
Seuthes arose with them, not in the least like a drunken man.

Perhaps in the two last mentioned uses the participle in the dative is better to be explained as a nominal attributive participle with no article (an equivalent to an object-noun/noun phrase).

NOTE 3: As in the case of the subject of the infinitive, raising-to-subject occurs with the participial clause used in indirect discourse, if governed by specific, otherwise impersonally used, verbs, such as ἀγγέλλεται "it is announced that", φαίνεται or φανερόν ἐστιν "it is apparent, manifest that", δῆλόν ἐστίν and ἔν-, κατά-δηλόν etc. ἐστιν "it is clear, obvious that" etc. Yet, with the participial clause the raising is not optional but obligatory. The difference can be seen when the use of the participial clause is compared/contrasted with the analogous use of an equivalent, semantically and syntactically, complementary ὅτι/ὡς clause (a "that" clause used as first argument, as the subject).

With a "that" clause:

τῷ δὲ Κλέωνι... ἀγγέλλεται... ὅτι ἥ τε στρατιὰ ἅπασα φανερά (verb ἐστι "is" is omitted) τῶν πολεμίων ἐν τῇ πόλει καὶ ὑπὸ τὰς πύλας ἵππων τε πόδες πολλοὶ καὶ ἀνθρώπων ὡς ἐξιόντων ὑποφαίνονται.
and it was told... to Cleon... that the whole army of the enemies was plainly to be discerned within the town, and that the feet of many men and horses, as if a sally to be intended, were visible under the gate.

With a participial clause (obligatory raising):

καὶ τῇ ὑστεραίᾳ μελλόντων αὐτῶν ἐπιπλεῖν τῇ Σάμῳ ἀγγέλλεται ὁ Στρομβιχίδης ταῖς ἀπὸ τοῦ Ἑλλησπόντου ναυσὶν ἀφιγμένος.
And the next day, when they meant to sail against Samos, it is announced that Strombichides with his galleys obtained from the Hellespontians had arrived.

(Here ἀγγέλλεται ὁ Στρομβίδης ἀφιγμένος = ἀγγέλλεται ὅτι ὁ Στρομβίδης ἀφῖκται)

Raising-to-subject is also employed with the infinitive (here the information announced is presented as a rumor and not as a fact):

νῦν δὲ ὁ Ἀσσύριος εἰς τὴν χώραν αὐτοῦ ἐμβαλεῖν ἀγγέλλεται.
And now comes the report that the Assyrian is going to invade his country.

Raising-to-subject can occur even with an "that" clause:

A "that" clause with no raising-to-subject construction:

ἐπεί γε μὴν δῆλον τοῖς Θηβαίοις ἐγένετο ὅτι ἐμβαλοῖεν οἱ Λακεδαιμόνιοι εἰς τὴν χώραν αὐτῶν, πρέσβεις ἔπεμψαν Ἀθήναζε λέγοντας τοιάδε.
But when it became clear to the Thebans that the Lacedaemonians were about to invade their land, they sent ambassadors to Athens to say the following.

A ""that" clause with raising-to-subject:

οἱ δὲ κατὰ στόμα δὴ (πολέμιοι, implied by context) ἔτι μόνοι ἐλύπουν καὶ δῆλοι ἦσαν ὅτι ἐπικείσονται ἐν τῇ ἐξόδῳ τε καὶ καταβάσει.
Οnly the enemies facing them in the front were still left to trouble them and it was manifest that they intended to attack them in their attempt to pass out and down the hill.

Here instead of:

δῆλον ἦν [ὅτι οἱ κατὰ στόμα πολέμιοι ἐπικείσονται...] (no subject raising, and "that" clause construed as a subject complement)

we have:

οἱ κατὰ στόμα πολέμιοι δῆλοι ἦσαν [ὅτι ἐπικείσονται...] (= δῆλοι ἦσαν ἐπικεισόμενοι, if with the participle) (subject raised from the embedded depended clause to the independed one, and "that clause" construed as an object complement)

and a possible reason for Xenophon choosing the raising construction is the coordination between ἐλύπουν and δῆλοι ἦσαν (a common subject).

NOTE 4: Cases of supplementary participles, yet not unquestionably, may also be considered to be participial phrases which shows raising of any sort:

(i) A kind of raising-to-object seems to occur with some participial phrase after prepositions. Even though these participles traditionally are not clearly classified as supplementary ones, the raising mechanism is the same and the syntactical function of the participle (complement of the preposition) may support such an interpretation. Some examples are the following:

καὶ οἱ Χαλκιδῆς ἐκ Νάξου ὁρμηθέντες ἔτει πέμπτῳ μετὰ Συρακούσας οἰκισθείσας Λεοντίνους τε πολέμῳ τοὺς Σικελοὺς ἐξελάσαντες οἰκίζουσι, καὶ μετ᾽ αὐτοὺς Κατάνην.
and the Chalcideans, going from Naxos, in the fifth year after the founding of Syracuse, built Leontium, expelling thence the Siculi, and after that Catana.

Semantically, in the temporal prepositional phrase "μετὰ Συρακούσας οἰκισθείσας" "after the founding of Syracuse" the complement of the preposition should better be considered to be the participle than the noun, because the prepositional phrase "(ἔτει πέμπτῳ) μετὰ Συρακούσας" ""(the fifth year) after Syracuse" cannot have any temporal rendering. So, here, the participial clause “Συρακοῦσαι οἰκισθεῖσαι” (here arbitrarily in nominative agreement) "Syracuse having been colonized" seems to underlie and as a whole function as the preposition’s complement; and since the specific preposition μετὰ requires a complement in the accusative, the participial clause is put into that case. But see also here.

The same approximately stands for the following:

ἡ ὀργὴ οὖν αὕτη… σὺν τῷ φόβῳ λήγοντι ἄπεισι.
This wrath, therefore,… will be gone with the passing/cessation of his fear.

Here, again, if the participle is left aside as an optional constituent of the prepositional phrase, the meaning of the phrase “συν τῷ φόβῳ” “together with his fear” is quite different than the original one where the participle is present and a temporal rendering is clearly seen.

(ii) Raising-to-subject could be also acknowledged in the following:

ἐλύπει αὐτὸν ἡ χώρα πορθουμένη.
Τhe ravaging of the country grieved him.

Here the noun "ἡ χώρα" "the country" is traditionally construed as the grammatical subject; but if the participle is taken away, the sentence "ἐλύπει αὐτὸν ἡ χώρα" "the country grieved him" is here almost meaningless, if not semantically ungrammatical.

Another argument in favor of this explanation can be the fact that in the above example the whole participial phrase is constructed as the first argument (the subject) of the verb ἐλύπει, while in other cases, like in the following, the participial phrase is constructed as the second argument (the object) of the verb, and for this kind of structuring no controversy between grammarians in textbooks exists, the participle of that sort being manifestly classified as a supplementary one:

ὡς ἐπύθοντο τῆς Πύλου κατειλημμένης, ἀνεχώρουν κατὰ τάχος ἐπ' οἴκου.
as soos as they were informed of the occupation of Pylos, they were in a harry to go back home.

For the traditional explanation of (some of) the above examples see for example here and here and here.[48]

The supplementary participle non in indirect discourse

When this participle is used not in indirect discourse, the following constructions are possible:

i) It is part of a predicate formed out by a copula verb, used mostly as an auxiliary verb; such verbs are: εἰμί “be”, γίγνομαι “become”, rarely ὑπάρχω “happen to be”.

προσεοικότες γίγνονται τοῖς γονεῦσιν οἱ παῖδες.
The children take after their parents.

ii) It forms a predicate with a quasi-copular verb that expresses a way of being, like δῆλός εἰμι, φαίνομαι/φανερός εἰμι be manifest, λανθάνω be hidden, unobserved, unseen, unnoticed, unaware etc., οἴχομαι be gone, away or absent, τυγχάνω happen to be, φθάνω be beforehand with etc. Semantically, with all these verbs it is the participle itself that expresses the main action, while in many cases the quasi-copular verb qualifies this action almost like an adverbial constituent.

ὁ δ' ἄρα ἐτύγχανεν ὢν εἰς φρόνησιν οὐδὲν βελτίων βατράχου γυρίνου, μὴ ὅτι ἄλλου του ἀνθρώπων.
He was after all in the very moment no better in intellect than a frog’s tadpole, to say nothing of being better than any other man.
δῆλος ἦν ἐπιθυμῶν προσελθεῖν.
It was evident that he wanted to come/ Evidently he wanted to come.

All these verbs always have a subject that is corefential with the subject of the participle; only the verbs λανθάνω and φθάνω may have an accusative object, denoting the person from whom someone/something is kept hidden or the person that someone overtakes in the action expressed by the participle.

ἔλαθεν αὐτὸν ἀπελθών.
He left without being noticed by him.
ὁ δὲ Λύσανδρος ἔφθη τὸν Παυσανίαν ἐν τῷ Ἁλιάρτῳ γενόμενος.
Lysander was beforehand with Pausanias in coming to Aliartus/ Lysander came to Aliartus before Pausanias.

NOTE: some of those verbs (such as verbs of presentation: δῆλός εἰμι, φανερός εἰμι, φαίνομαι etc.) may also be used as governing verbs in indirect discourse clausal constructions introduced by the particles ὅτι or ὡς.

iii) It is the complement of verbs denoting ‘’’commencement, continuation (patience, tolerance) or termination (fatigue) of an action’’’, and it is always in the present tense stem; such verbs are: ὑπάρχω begin, take the initiative in, ἄρχομαι start, begin, παύω cause to cease, stop from, παύομαι cease, stop, λήγω leave off, cease, ἀπαγορεύω give up, κάμνω be tired or weary, διαβιῶ spend my whole life, διαμένω, διάγω, διαγίγνομαι, oὐ διαλείπω, διατελῶ continue, keep up, etc., ἀνέχομαι tolerate, περιοράω overlook etc. With these verbs the participle’s subject is coreferent with the verb’s subject, if there is no object in the structure, or else with the object, and the participle agrees in case (nominative or accusative) with this word.

ἄρξομαι διδάσκων περὶ τοῦ θείου.
I will start teaching about god.
οὐκ ἀνέχομαί σε ὑβρίζοντα.
I don't tolerate you insulting me.
ἔπαυσεν δὲ τοὺς συμμάχους ὑμῶν ἀφισταμένους.
He stopped your allies from revolting from you.

iv) A supplementary participle can be used with verbs expressing passions of the soul, such as ἀγανακτῶ be vexed, αἰσχύνομαι be ashamed, ἥδομαι be pleased, χαίρω be happy, λυποῦμαι be sorry, ὀργίζομαι get angry etc.

χαίρω ὁρῶν σε ὑγιῆ.
I am glad to see you healthy.
ἀγανακτεῖ ὑπομένων τὰς ὕβρεις.
He is vexed tolerating the insults.

v) Finally, it is used with some verbs not easy to classify: εὖ/καλῶς/κακῶς/δίκαια/δεινὰ ποιῶ behave well, honourably, fairly, badly, χαρίζομαι gratify, favour, ἀδικῶ do wrong, νικῶ defeat, περιγίγνομαι prevail, survive, κρατῶ have the better of, ἡττῶμαι be worsted by, λείπομαι be left behind, fall short.

καλῶς ποιεῖτε τοὺς γέροντας ἐπιμελοῦντες.
You are doing well by taking care of the old men.
The circumstantial (adverbial) particple

The circumstantial participle, used as a satellite of another verbal form, is always without the article (i.e. it is put in the predicative position). It is added to a noun or pronoun to denote the circumstance under which the action of another verbal form (a finite verb or an infinitive/another participle) takes place. The action of the main verb is the main one.

It is also called adverbial because it qualifies the main verb like any other adverb, adverbially used adjective, adverbial prepositional phrase, adverbial clause or supplementary predicate. In most cases it has the force of a dependent clause denoting time, cause, purpose, supposition, opposition, concession. Often it denotes manner-means or any other attendant circumstance.

Two main constructions can be distinguished:

i) the participle agrees in case (and most of the times in gender and number) with a noun or pronoun that is an argument of the main verb, usually subject, direct or indirect object or dative of interest of any kind. In this case the subject of the participle is coreferent to that verbal argument (participium conjuctum). ii) participial phrases, composed by the participle and a subject, which form a full new predicate, additional to the verb predicate.

Two subtypes can be distinguished:

(1) Genitive absolute: the participle and a noun or pronoun (it’s subject) stand in the genitive case; in this construction normally the subject of the participle has no coreference to any other verbal argument, especially the subject. (2) Accusative absolute: it is found when the verb in which the participle belongs is impersonal or so used, or when it is an impersonal expression; in this case the subject of the participle is usually an infinitive, as it would be if an impersonal verb in some finite mood was used. The participle is always in the neuter gender. Nevertheless, it is possible for a personal verb’s participle to stand with its subject in the accusative absolute construction, if only it is preceded by the particles ὡς or ὥσπερ and expresses cause or conditional comparison respectively.

USE OF TENSES: Normally all circumstantial participles, according to the grammatical tense stem used for their morphological formation, express only time relative to that of the main verb, but they always express stage of action. Nevertheless, the future stem is only used for denoting purpose, and seldom for denoting future cause (in the latter case normally the particle ὡς precedes the participle).

Kinds of circumstantial participles:

  • The temporal participle is used in the present or in the aorist tense stem (the perfect being rare enough). It expresses a simultaneous or an anterior action. It is usually found with temporal adverbs such as ἅμα while, simultaneously, ἐνταῦθα then, ἔπειτα after, εὐθύς immediately, ἤδη already, μεταξύ meanwhile.
ἀποπλεύσας εἰς Λάμψακον τὰς ναῦς ἐπεσκεύαζεν.
After having sailed to Lampsacus he repaired the ships.
ἐπαιάνιζον ἅμα πλέοντες.
They were singing the paean while they were sailing.
  • The causal participle is used with every tense stem (rarely future when preceded by the particle ὡς).
ὁρῶν αὐτοὺς λυπουμένους ὑπεσχόμην γράψειν τὴν ἐπιστολήν.
As I saw that they were sad I promised to write the letter.
εἰδώς σε ἱκανὸν ὄντα οὐ φοβοῦμαι.
As I know that you are competent, I am not afraid.
Some particles may precede a causal participle:
a. ἅτε, οἷον, οἷα, ἅτε δή, οἷον δή, οἷα δή: the reason/cause is presented by the speaker, narrator, writer etc. as an independent fact (objective reason/cause):
ἅτε δὲ ἀήθους τοῖς Λακεδαιμονίοις γεγενημένης τῆς τοιαύτης συμφορᾶς, πολὺ πένθος ἦν κατὰ τὸ Λακωνικὸν στράτευμα.
As in fact such a calamity had been unusual with the Lacedaemonians, there was great mourning throughout the Laconian army.
b. ὡς: the reason/cause is presented as an idea, thought or personal opinion etc. of the subject of the main verb (subjective cause or reason).
Τὸν Περικλέα ἐν αἰτίᾳ εἶχον ὡς πείσαντα σφᾶς πολεμεῖν.
They found fault with Pericles, on the belief/ground that he had persuaded them to engage in the war.
  • The final (telic) participle (expresses purpose) is used with the future tense stem. It forms the negation with μή. If the participle is modifying a verb that expresses movement, then it usually stands alone. If the verb does not express a movement then the participle is often found with the particle ὡς (in this case the intention of the subject is underlined as a personal consideration, and in many cases it is difficult to determine whether this participle is final or causal).
ἀνεχώρησεν ἀπαγγελῶν τὰ γεγονότα
He left in order to announce the events.
ψεύδεται ὡς κρύψων τὴν ἀλήθειαν.
He lies in order to hide the truth. (or causal: on the belief that he will hide the truth)
ἔπεμψεν Ἀριστοτέλη ἀγγελοῦντα τὰς σπονδάς.
He sent Aristotle in order for him to announce the agreements (Aristotle will announce).
  • The conditional participle is used in all tense stems except the future (negated by μή). It stands as the protasis (:hypothesis) of a conditional sentence, the main verb being the apodosis. It can express any type of conditional thought, but is by far more often used alongside with potential moods or future indicative (and future-like expressions), expressing any kind of future condition.
ὅ νῦν ὑμεῖς μὴ πειθόμενοι ἡμῖν πάθοιτε ἄν (= εἰ μὴ πείθοισθε).
That might well happen to you now if you do not listen to us.
  • The concessive participle (denotes opposition, concession, limitation etc.) may be preceded by the particles καὶ, καίπερ, καὶ ταῦτα, οὐδέ, μηδἐ (= although) or/and followed by ὅμως (= nevertheless) in the main verb structure.
Ἀγησίλαος δέ, καίπερ αἰσθανόμενος ταῦτα, ὅμως ἐπέμενε ταῖς σπονδαῖς.
Agesilaus, although he was aware of those things, nevethrless he continued to be steadfast in the truce.
  • A participle may also express any other attendant circumstance under which an action takes place:
παραλαβόντες τοὺς Βοιωτοὺς ἐστράτευσαν ἐπὶ Φάρσαλον.
Having taken the Boeotians with them, they marched against Pharsalus.


The gerundive is a passive verbal adjective that indicates the necessity for the action of the verb to be performed. It takes the nominative endings -τέος, -τέᾱ, -τέον, declining like a normal first/second declension adjective. Its stem is normally of the same form as the aorist passive, but with φ changed to π and χ to κ, e.g.

παύω → παυστέος (to be stopped)
λαμβάνω → ληπτέος (to be taken)

Gerundives may be used as straightforward adjectives, with the agent, if any, in the dative:

βοῦς θυστέος ἐστίν
An ox must be sacrificed

They may also be used to express impersonal necessity

ποιητέον (ἐστί) ...
It is necessary to do...

Time and aspect

One of the most notable features that Ancient Greek has inherited from Proto-Indo-European is its use of verb "tense" to express both tense proper (present, past, or future) and the aspect of the time (as ongoing, simply taking place, or completed with a lasting result). The aspectual relation is expressed by the tenses in all the moods, while the temporal relation is only expressed in the indicative and to a more limited extent in the other moods (also called the dependent moods).

With regard to the time relation that they express in the indicative, the seven tense-aspects are divided into two categories:

This classification, which properly applies only to forms of the indicative, is also extended to the dependent moods in the cases where they express the same time relation as the indicative. The time relation expressed by a verb's tense may be present, past or future with reference to the time of the utterance or with reference to the time of another verb with which the verb in question is connected. Compare for instance ἀληθές ἐστιν "it's true" with εἶπον ὅτι ἀληθὲς εἴη "I said that it was true" or "I said 'it's true'."

A verb also expresses one of three possible aspects, irrespective of the mood it may be in:

  • Imperfective aspect: indicating an ongoing, continuous, or repeated action. The present and the imperfect convey this aspect.
  • Perfective aspect (traditionally also called aorist aspect in Greek grammar): indicating that the action is started and concluded at the same time, or that the action is focused on a single point in time, or that the action simply occurs without reference to its duration or lasting effect. The aorist conveys this aspect in all moods.
  • Perfect (traditionally also often called perfective, but not to be confused with the above): indicating that the action is completed with a result that remains into the time being considered. The perfect (in all moods) as well as the pluperfect and future perfect carry this combination of relative tense and aspect.

Mood of the dependent verb

The rules on mood sequence (consecutio modorum) determine the mood of verbs in subordinate clauses in a way analogous to but more flexible than the Latin rules on time sequence (consecutio temporum) that determine their tense.

Putting aside special cases and exceptions, these rules can be formulated as follows:

  • In dependent sentences, where the construction allows both the subjunctive and the optative, the subjunctive is used if the leading verb is primary, and the optative if it is secondary. E.g. πράττουσιν ἃ ἂν βούλωνται, "they do whatever they want"; but ἐπραττον ἃ βούλοιντο, "they did whatever they wanted".
  • Similarly, where the construction allows both the indicative and the optative, the indicative follows primary, and the optative follows secondary tenses. E.g. λέγουσιν ὅτι τοῦτο βούλονται, "they say they want this"; εἶπον ὅτι τοῦτο βούλοιντο, "they said they wanted this".

See also


  1. BRANDÃO, Jacynto L.; SARAIVA, Maria O. de Q.; and LAGE, Celina F. Ελληνικά: introdução ao grego antigo. Belo Horizonte (Brazil): Editora UFMG, 2005. p. 44, 67 and 512.
  2. 2.0 2.1 FREIRE, Antônio. Gramática Grega. São Paulo (Brazil): Martins Fontes, 1987. p. 17.
  3. Rijksbaron, Albert, The syntax and semantics of the verb in classical Greek, The University of Chicago Press, 2006, pp. 95-115.
  4. Kühner, Raphael. Grammar of the Greek language for the use in high schools and colleges. (Translated by B. B. Edwards and S. H. Taylor). 1844. Pp. 449 ff.
  5. See for example from here and the §§ff.
  6. See for example here.
  7. See for example here.
  8. See for example here.
  9. Madvig, J.N., Syntax der griechishen Sprache, besonders der attishen Sprachform, für Shulen. Braunsweig 1847, pp. 187ff.
  10. Rijksbaron, Albert. The syntax and semantics of the verb in classical Greek. The University of Chicago Press, 2006, pp. 96ff., §§31ff.
  11. Smyth, Herbert Weir, Greek Grammar, Harvard University Press, 1984 (renewed), pp.442ff.,§§ 1988ff. Henceforth, quotations given from the online version: §§ 1988ff..
  12. William Watson Goodwin, Syntax of the Moods and Tenses of the Greek Verb. Cambridge University Press, Third edition, 1867, p.12 § 15; p. 22 § 18.2.b; p. 30 § 23.1; p. 212 § 106, REMARK. Henceforth quotations from the online version: § 96ff. § 87.
  13. William Watson Goodwin, Syntax of the Moods and Tenses of the Greek Verb § 73.
  14. Rijksbaron, Albert. The syntax and semantics of the verb in classical Greek. University of Chicago Press, 2006, pp. 102-103, §33.1, and pp. 44–45, §16.2.
  15. Rijksbaron, Albert. The syntax and semantics of the verb in classical Greek. University of Chicago Press, 2006, pp. 106ff., §33.2.
  16. William Watson Goodwin. Syntax of the moods and tenses of the Greek verb § 115.
  17. Herbert Weir Smyth. A Greek Grammar for Colleges, § 2019 and § 2023.
  18. Rijksbaron, Albert. The syntax and semantics of the verb in classical Greek. The University of Chicago Press, 2006, p. 109, §33.1, and pp. 44–45, §16.2.
  19. Herbert Weir Smyth, A Greek Grammar for Colleges §1998, § 1999, § 2024
  20. Herbert Weir Smyth, "A Greek grammar for colleges"§ 2257-2259.
  21. Herbert Weir Smyth, "A Greek grammar for colleges"§§ 2453-2461.
  22. Herbert Weir Smyth, "A Greek grammar for colleges"§ 2497.
  23. Herbert Weir Smyth, "A Greek grammar for colleges"§ 2279.
  24. For examples see here: Willam Watson Goodwin, Syntax of the Moods and Tenses of the Greek Verb § 755.
  25. Herbert Weir Smyth, A Greek grammar for colleges §§ 1972-1981.
  26. Kühner, Raphael. Grammar of the Greek language for the use in high schools and colleges. (Translated by B. B. Edwards and S. H. Taylor). 1844. P. 455
  27. Kühner, Raphael. Grammar of the Greek language for the use in high schools and colleges. (Translated by B. B. Edwards and S. H. Taylor). 1844. Pp. 454-455: "When the subject of the governing verb... is at the same time the subject of the infinitive also, the subject is not expressed by the acc. of a personal promoun in Greek, as in Latin, but is wholly omitted, and when adjectives and substantives stand with the infinitive, to explain or define the predicate, they are put, by attraction, in the nominative".
  28. Kühner, Raphael. Grammar of the Greek language for the use in high schools and colleges. (Translated by B. B. Edwards and S. H. Taylor), 1844, p. 456, remark 6: The Greeks are fond of changing the impersonal construction into the personal, by elevating the object into a subject, making the impersonal expression, personal, and referring to the person of the subject.
  29. See for example here:H. W. Smyth A Greek grammar for colleges, §§ 1982-1983.
  30. See here., for example, a translation of the last passage.
  31. Rijksbaron, Albert. The syntax and semantics of the verb in classical Greek. The University of Chicago Press, 2006, p. 99 § 32.1.
  32. Kühner, Raphael. Grammar of the Greek language for the use in high schools and colleges. (Translated by B. B. Edwards and S. H. Taylor). 1844. Pp. 453-454.
  33. Herbert Weir Smyth,A Greek Grammar for Colleges, §§ 1060 ff.
  34. Herbert Weir Smyth, A Greek Grammar for Colleges, § 1973, c.
  35. Syntax of the Moods and Tenses of the Greek Verb, by William Watson Goodwin, §§ 138 ff.
  36. Syntax of the Moods and Tenses of the Greek Verb, by William Watson Goodwin §§ 187 ff.
  37. Rijksbaron, Albert, The syntax and semantics of the verb in classical Greek, The University of Chicago Press, 2006, p. 95 and pp. 116 ff.
  38. Syntax of the Moods and Tenses of the Greek Verb, by William Watson Goodwin §§ 821 ff.
  39. William Watson Goodwin, Syntax of the Moods and Tenses of the Greek Verb, § 827.
  40. A Greek Grammar for Colleges, by Herbert Weir Smyth, §§ 1122-1124
  41. A Greek Grammar for Colleges, by Herbert Weir Smyth §§ 1119-1120.
  42. A Greek Grammar for Colleges, by Herbert Weir Smyth, § 2052
  43. A Greek Grammar for Colleges, by Herbert Weir Smyth §§ 2049-2053.
  44. A Greek Grammar for Colleges, by Herbert Weir Smyth, §§ 1168 ff.
  45. Rijksbaron, Albert. The syntax and semantics of the verb in classical Greek. University of Chicago Press, 2006, p. 118, note 2.
  46. Rijksbaron, Albert. The syntax and semantics of the verb in classical Greek. The University of Chicago Press, 2006, p. 118, Note 1
  47. See for example: Ilja Seržant, An approach to syntactic reconstruction, p. 139. In: Perspectives on Historical Syntax, Studies in language Companion 169, 2015, John Benjamins Publishing Company
  48. Kühner, Raphael. Grammar of the Greek language for the use in high schools and colleges. (Translated by B. B. Edwards and S. H. Taylor). 1844. P. 470, REMARK No 5.

External links

de:Altgriechische Grammatik

fr:Grec ancien