Greek Sentence Structure: Loose and Periodic Style


Introductory Note

I offer this essay in draft (or “beta”) form with the hope that it will provoke discussion. Corrections and criticisms are welcome. Just e-mail me by clicking here.

Graphic images of the Greek texts discussed are provided. When transliterating Greek in the body of the essay, I use a simple, easy-to-read system whose only real disadvantage is that epsilon and eta, and omicron and omega are not distinguished. With the Greek text at hand, this should cause no problem.

—Hardy Hansen

Technical Terms: What Are They For?

veni, vidi, vici. These words, which Suetonius tells us were carried on a banner in a triumphal procession of Julius Caesar (Divus Iulius 37), can serve as a starting-point for an analysis of rhetorical figures and sentence-construction. “I came, I saw, I conquered.” Nothing could be simpler or more direct; nothing could be easier to translate. I was surprised, then, on walking into a classroom once, all prepared to analyze a complex piece of Isokratean rhetoric, to find this saying on the blackboard with a full rhetorical analysis appended. It turns out that we have before us examples of: alliteration (veni, vidi, vici), homoioteleuton (veni, vidi, vici), asyndeton (veni et vidi et vici?), trikolon (how about veni, vici?), isokolon (try adding a syllable: veni, vidi, victi sunt), composition in short kommata instead of longer kola (compare: ad hostes adveni et, postquam illorum copias vidi, cunctos facile vici), spondaic rhythm, paromoiosis (similar structure of kola or kommata), parechesis (similarity of sounds, especially between vidi and vici), and perhaps paronomasia (a play on words involving, again, vidi and vici), to say nothing of the overall brevity achieved by the ellipsis of words easily supplied from the (very crowded) context. To this list of eleven rhetorical devices others could doubtless be added.

This analysis is a joke, of course. It pokes fun at pedantry. But it is also perfectly valid and makes a serious point. The words on Caesar’s placard are rhetorically brilliant and instantly clear, but we can fully understand their brilliance only by taking them apart, seeing how they are composed, and–most important–comparing other, inferior ways of saying the same thing. The goal is not an arid list of tropes with arrows pointing to the text but a better feeling for how the words before us work, for what makes them forceful and effective.

How big an armory of rhetorical terms does a student of Greek prose need? Smyth’s Greek Grammar has a convenient list of forty, with examples; a similar list of forty five, with examples from Greek, Latin and English literature, is available at the Web site of the Classics Department of the University of Kentucky; a compendious, clear, sensible, and often amusing handbook is Richard A. Lanham‘s A Handlist of Rhetorical Terms. Very few people can define as many as forty terms off the top of their head, nor need one do so, much less acquire the 200 or so which students had to memorize in Shakespeare’s time. Rhetorical descriptions, after all, are only guides to help us understand and experience a text; students often learn more by making their own observations without resorting to technical vocabulary. Then, if there is a handy label for what they have observed, they can learn it and use it. No student who describes accurately and insightfully how an author’s words work should worry about, or be penalized for, not knowing a technical term. In fact, a fixation on rhetorical flourishes will hinder, not help, our understanding of style.

Consider again Caesar’s placard. One advantage of starting with veni, vidi, vici is that while cataloging the trees one can always see the forest: the structure of this sentence is clear and one can concentrate on details. When analyzing complex sentences, however, it is easy to forget that the goal of arranging words “rhetorically” is to shape a sentence which develops clearly as it is heard or read, not to distract the audience with pointless ornamentation. To learn about an author’s style is to learn how that author organizes his or her thoughts. Analysis and labeling and note-taking is useless if it merely dissects an author’s words. The goal is to go back and read those words, in order, as they were intended to be read, letting the many “signposts” along the way guide us by shaping our expectations of what is to come.

This point cannot be emphasized enough. Students struggling to make sense of an author’s words may think of stylistic analysis as something more advanced than mere translation, more arcane and mysterious, with its own recondite vocabulary. Ironically, they get this impression from the emphasis which most of us place on the most obvious marks of style–rhetorical figures such as chiasmus, alliteration, homoioteleuton–as the key to everything: these strike our ears and eyes as we read, so naturally we begin our analysis by scanning the text, collecting as many tropes as we can, and labeling them like specimens found on a nature hike. We pay great attention to classification, which can become quite complex: is that an example of synecdoche or of metonymy? Are we dealing with paronomasia or only parechesis?

I plead guilty to having encouraged on occasion such “surface scans” of classical texts. The problem with them is that they don’t necessarily lead students to understand how Gorgias or Demosthenes or Isokrates has structured and expressed his thoughts, the “character” or “cast of thought” each presents to us. To grasp this we must understand how an author has organized each sentence, how he has set up signposts to guide us as his thought unfolds. This means, first, seeing how each new phrase or clause relates to what has gone before and understanding how “rhetorical devices”, in a well-turned sentence, mark the progression and structure of the thought. The real rhetoric is in the unfolding of the ideas in the order in which the author wants us to hear or read them; rhetorical tropes are markers along the way. To learn about an author’s style is to learn to read that author’s words, in order, as they were written–not to disarrange the words but to hear what the ancient audience heard.

In his admirable essay on the style of Cicero W. Ralph Johnson puts it succinctly: ” . . . Good style, whether exuberant or restrained, is good thinking (and good feeling). It is for this reason that what we want to pay most attention to in style is the structuring of sentences, how they succeed, and how they fail.” [Luxuriance and Economy: Cicero and the Alien Style, p. 7] He quotes Virginia Woolf’s description of “the rhythm of a book that, by running in the head, winds one into a ball”, and he rescues from obscurity a book by Edith Rickert, New Methods for the Study of Literature, which asserts the importance for style of the length of whole sentences and of their parts. [Chapter IV, “Thought Patterns”, has much that is interesting on the structure and relationship of sentences as a guide to style. As Rickert remarks (p. 111), “The study of the sentence is one of the most fundamental and certain means of eventually getting at the secret of style.” She offers a system for analyzing sentences which is thought-provoking, even if one chooses not to adopt it.]

[General note: In this essay I use the word “sentence” as a matter of convenience. There is no exactly corresponding Greek term or concept. Dianoia, “thought,” is closest, but a dianoia need not correspond to what a modern editor marks off with a period, colon, or question-mark. Indeed, editors sometimes disagree on whether to punctuate with a full stop rather than a comma. See now Dover, The Evolution of Greek Prose Style, chapter 2.]

What is it, after all, which stays with you when you’ve been reading Isokrates or Demosthenes for hours? Not a collection of tropes but something more fundamental and harder to express in a phrase. Why did reading Isokrates for hours on end once make me physically ill? Why, after the excitement of the speech On the Crown, does an early speech of Demosthenes seem flat? Because a real stylist “gets in among” you. The “rhythm” of the writer’s thought is what does this.

The Architecture of Greek Sentences: Loose and Periodic Style

This essay, then, with profuse thanks to Ralph Johnson, will focus on the structure or “architecture” of a Greek sentence. A formidable phrase, perhaps, but less so if one bears in mind two fundamental points:

1. Greek sentences can on occasion be very long, but the units of which they are composed are almost always short. “The edifice, lofty though it may be, is built of bricks, not of huge blocks of Cyclopean masonry.” [J. D. Denniston, Greek Prose Style, p. 61]

2. Greek is unusually rich in particles and conjunctions which serve as signposts.

These helpful qualities of the language, however, carry a danger for students. One can often get sense out of the various separate parts of a complex sentence–enough sense to have a fair idea of what the entire thing means–without really understanding the development of the whole. And the reason for this is often the very profusion of “marker-words” or “signposts” in Greek: because they are omnipresent, students tend to ignore them or at least to forget earlier signposts when later ones, in corresponsion with them, appear. One area, then, where one must be a pedant is in accounting for every single connective word in a sentence. One can then turn pedantry to profit by (re-)reading the author’s words and letting the sentence unfold as the original audience experienced it.

Before we turn to the analysis of sample passages from Greek authors, let us consider the two main types of sentence structure which ancient critics recognized: the loose style (lexis eiromene) and the periodic style (lexis katestrammene). The first phrase means literally “speech strung together” (from eiro, “to string or thread together”, like beads in a necklace); the second, “speech turned or guided toward an end”; the word “period” (periodos, “way around”) refers metaphorically to a racecourse, where the starting and finish lines were the same: contestants went out and around the turning-post, then retraced their path. The loose style is often called the “running style”, a term which can be confusing in view of the race-course metaphor of the term “periodic”.

[Dover points out the difficulty of interpreting the terms lexis katestrammene and periodos (The Evolution of Greek Prose Style, chapter 2, Appendix: Period and Kolon). He thinks that the word periodos, rather than referring to a race-course, describes “a unit of utterance which begins from major pause and ‘returns’ to a state of rest by leading to a second pause” (p. 39).]

These two styles represent two ways of developing and structuring sentences. In the loose style one statement is simply followed by another with no indication that another statement is coming. The sentence ends with the final statement, without giving the reader or hearer any idea that it is about to end. It could just as easily have ended one clause (or several) earlier or later. Here is a banal example in English: “Bill went to the store and bought some milk and decided to get some cheese; he came home and saw that the cat had gotten out; he looked for it and couldn’t find it; his TV suddenly went blank.” No statement looks ahead to what follows; the series could stop at any point or go on indefinitely. There is no “race-course” with a beginning, a turning point, and an end. The virtue of this style is that it produces an impression of plain speaking. In a court case, for example, if one is trying to set out the facts and convince the jury of one’s honesty and straightforwardness (no matter how one has actually slanted the “facts”), this is the style to use. Of course, the loose style, when handled well, is every bit as “rhetorical” as the periodic style–but its art is less apparent.

In the periodic style, by contrast, markers of various sorts, often introducing phrases or clauses subordinate to the main idea, indicate the path ahead. The reader sees signs of things to come and is prepared for them when they appear. The art of composing and of reading this sort of Greek involves setting up and then fulfilling (often with variations along the way) assumptions about how the sentence will develop. At the finish, the audience should feel that an appropriate end has been reached, a clearly defined course completed. All periodic composition involves, in one way or another, suspension of sense: the arrangement of one or more elements of the sentence so that the thought is not felt to be complete until something else has been added.

Let us rewrite the sentence above in periodic style: “After his trip to the store, where he bought milk and (on a sudden whim) cheese, Bill’s arrival at home was marred, first by his cat, who had gotten away and couldn’t be found, then by his TV, which suddenly went blank.” Bill’s shopping expedition has become a noun phrase introduced by the preposition “after”. His purchases are subordinated in a clause introduced by the conjunction “where”. A parenthesis characterizes his acquisition of cheese. The center of the sentence makes the main point: coming home was no fun. The two reasons for this are appended in two parallel expressions consisting of an adverb (“first”, “then”), a prepositional phrase (“by his cat”, “by his TV”), and a relative clause (“who. . .”, “which. . .”).

As silly as these sentences are, they illustrate an important point: unlike a simple narration in the running style, a periodic sentence will highlight the main point(s) the writer is trying to put across and will require the writer to decide how various events or ideas are related. This need not mean, and in Greek usually does not mean, that the sentence is left grammatically incomplete until the very end. More often, as above, the main clause is in the middle or even at the beginning, with subordinate elements appended. Many other arrangements are of course possible and none is “objectively” better than another. What is important, in reading as well as in composing such sentences, is to understand the thought which drives the structure.

Loose Style and Periodic Style: Examples in English

The periodic style is not at home in present-day American English. Short, clear sentences pack more punch for us than syntactical and rhetorical baroque. Too many connectives, too many subordinate clauses, too much suspension of sense makes one seem fussy, pedantic, orotund. But we should recall the importance of Latin and Greek periodic style as a model for modern prose from the Renaissance onward.

Here are examples in English of both styles. First, the lexis eiromene which has enjoyed such a vogue in English through the twentieth century, in a paragraph from Ernest Hemingway’s The Sun Also Rises:

The bus climbed steadily up the road. The country was barren and rocks stuck up through the clay. There was no grass beside the road. Looking back we could see the country spread out below. Far back the fields were squares of green and brown on the hillsides. Making the horizon were the brown mountains. They were strangely shaped. As we climbed higher the horizon kept changing. As the bus ground slowly up the road we could see other mountains coming up in the south. Then the road came over the crest, flattened out, and went into a forest. It was a forest of cork oaks, and the sun came through the trees in patches, and there were cattle grazing back in the trees. We went through the forest and the road came out and turned along a rise of land, and out ahead of us was a rolling green plain, with dark mountains beyond it. These were not like the brown, heat-baked mountains we had left behind. These were wooded and there were clouds coming down from them. The green plain stretched off. It was cut by fences and the white of the road showed through the trunks of a double line of trees that crossed the plain toward the north. As we came to the edge of the rise we saw the red roofs and white houses of Burguete ahead strung out on the plain, and away off on the shoulder of the first dark mountain was the gray metal-sheathed roof of the monastery of Roncesvalles.

(Ernest Hemingway, The Sun Also Rises, New York 1926, repr. 1954, chapter 3, p. 108)

As an example of periodic style here is Samuel Johnson in the Preface to his edition of Shakespeare:

That praises are without reason lavished on the dead, and that the honours due only to excellence are paid to antiquity, is a complaint likely to be always continued by those who, being able to add nothing to truth, hope for eminence from the heresies of paradox, or those who, being forced by disappointment upon consolatory expedients, are willing to hope from posterity what the present age refuses, and flatter themselves that the regard which is yet denied by envy will be at last bestowed by time. . . .

To works, however, of which the excellence is not absolute and definite, but gradual and comparative; to works not raised upon principles demonstrative and scientific, but appealing wholly to observation and experience, no other test can be applied than length of duration and continuance of esteem. What mankind have long possessed they have often examined and compared, and if they persist to value the possession it is because frequent comparisons have confirmed opinion in its favour. As among the works of nature no man can properly call a river deep or a mountain high without the knowledge of many mountains and many rivers, so, in the productions of genius, nothing can be styled excellent till it has been compared with other works of the same kind. . . .

The reverence due to writings that have long subsisted arises, therefore, not from any credulous confidence in the superior wisdom of past ages, or gloomy persuasion of the degeneracy of mankind, but is the consequence of acknowledged and indubitable positions, that what has been longest known has been most considered, and what is most considered is best understood.

(Samuel Johnson, Selected Writings, ed. R. T. Davies, Evanston 1965, pp. 262-263)

The elegant and varied architecture of these sentences, made clear by a fullness of expression where parallel phrases reinforce one another, will serve to remind us that English, too, is capable not only of “straight talk” but also of grandiloquence. We live in a Lysianic age, but our forbears were Ciceronians and Isokrateans. [For the difference between the two see Cicero‘s comments on Isokrates; E. Laughton, “Cicero and the Greek Orators”; G. Williamson, The Senecan Amble (on stylistic models in the Renaissance).] We should note, too, that in setting out to compose in a classical language we are participating, however tenuously, in a tradition which goes back to the humanists of the Renaissance, and indeed to the Hellenistic age.

[Lanham (s.v. “period”, p. 113) quotes Michael Baxandall, Giotto and the Orators (Oxford 1971), pp. 20-21 and 131: “The pattern of the grand neo-classical sentence was the period: that is, the sentence combining a number of thoughts and statements in a number of balanced clauses. . . . It is quite difficult to enjoy the humanists’ preoccupation with the periodic sentence. . . . But one cannot come to historical terms with the humanists’ verbal performance without recognizing how supremely important it was for them, and in how many different ways. The periodic sentence is the basic art form of the early humanists. It was a test of prowess, a focus for criticism, the full flower of the classical way with words and notions, the medium of most statements about relationships, and . . . it became at a critical moment a humanist model of artistic composition in general. . . . Compositio was a technical concept every schoolboy in a humanist school had been taught to apply to language. It did not mean what we mean by literary composition, but rather the putting together of the single evolved sentence or period, this being done within the framework of a four-level hierarchy of elements: words go to make up phrases, phrases to make clauses, clauses to make sentences.”]

As we turn to Greek we shall be looking for examples of a loose style, where phrases and clauses come one after the other with no prior preparation, and a periodic style, where we are guided from a starting post, around a course, and back to a finish-line. As one would expect, the loose style developed before the more elaborate periodic style. But only scraps of the earliest prose remain, and the greatest writer associated with this style, Herodotos, is far more sophisticated than some caricatures of him suggest and writes just as elegantly and effortlessly in the periodic as in the loose style.

Loose or Running Style: Hekataios

The longest surviving fragment of Herodotos’ predecessor Hekataios (FHG 1, F15) provides a rare sample of a loose, naive style:

For a graphic image of the Greek text click here.

Orestheus, son of Deukalion, came to Aitolia for a kingdom and his dog gave birth to a log, and he ordered it to be buried, and from it grew a vine with many grapes. Thus he also named his own son Phytios. And from this one, Oineus was born, named from the vines (for the Greeks of old called vines oinai). And from Oineus was born Aitolos.

Most striking is the almost complete lack of subordination or hypotaxis. Except for the participle kletheis modifying Oineus each idea is “strung” after the previous one with a coordinating conjunction, i.e. paratactically (with coordination or parataxis). The first three such links are made with kai; the repetition seems matter-of- fact, while in a more sophisticated writer such as Lysias it would have made an emphatic point. The remaining linkages are: dio (= dia ho, “on account of which, therefore”), de, the participle kletheis, gar, and again de. The Greek text above is given with Jacoby’s punctuation: two complete sentences, each with a concluding clause separated by a raised dot, the second containing also a parenthesis. There is nothing, however, in either the grammar or the sense to prevent putting a full stop at the end of each clause, including the one set off by parentheses. (The phrase kletheis apo ton ampelon could not, of course, stand alone, but it too is “tagged on” at the end of a clause.) Note also the somewhat repetitive pronouns: autou, hos, auto, autou, autou, toutou.

Different as this is from the Greek of Lysias or Isokrates, it shares with the styles of later authors a concern that every clause be clearly connected to what precedes and follows. Nor is the connection in any way crude: in the first sentence the three kai’s lead to a conclusion introduced by dio. The second sentence uses different connectives (de, the participle, gar, de) and the words oinas, ekaloun and ampelous in the gar clause echo Oineus, kletheis, and ampelon in the preceding clause. Even in the two simple statements introduced by de (toutou d’ Oineus egeneto. . .Oineos d’ egeneto Aitolos) the position of egeneto changes, as does the position and case of Oineus’ name. This is not to claim any great sophistication for Hekataios’ Greek, but only to remind us that naivete is a relative notion.

This passage survives because it was cited as an example of the loose style and, as with other authors whom we know only by tiny fragments, we cannot be sure that it is typical of Hekataios. He may have assumed a more naive style when he was recounting a myth. Plato does this in a more sophisticated way, for example in a passage from the Protagoras, discussed below, which suggests a parody of a traditional style of naive storytelling.


Hekataios’ successor Herodotos is vastly more sophisticated. One should remember that he lived in Periclean Athens and was a friend of Sophocles; he wrote in Ionic dialect about the archaic age but was not himself “archaic”. He is a sophisticated storyteller (and historian!) who often employs, but is not limited to, lexis eiromene. Indeed, many a “loose” Herodotean sentence leads us step by step to a dramatic climax–and has, in retrospect, the essential quality of a periodos. Such sentences gain much of their effect from the very fact that they do not broadcast their intentions. Here is Herodotos’ description of the arrival of Adrastos at Sardis (1.35):

For a graphic image of the Greek text, click here.

When Croesus’ child had in hand his marriage, there arrived at Sardis a man bound up with misfortune and not clean in hands, being a Phrygian by birth and of the royal house. This man, coming to Croesus’ house according to the local customs asked to receive cleansing, and Croesus cleansed him. Cleansing is similar for the Lydians and the Greeks. When Croesus had performed the accustomed rites, he inquired where the man was from and who he was, saying as follows: “Man, who are you and from where in Phrygia have you come to be at my hearth? Whom of men or women did you kill?” And he answered: “King, of Gordias son of Midas I am the child, and I am called Adrastos, and having killed my brother unwilling I am here, driven out by my father and deprived of everything.” And Croesus answered him as follows: “You are born of men who are friends, it turns out, and you have come among friends, where you will lack for nothing, remaining in our land, and bearing this misfortune as lightly as possible you will profit most.”

[This translation is as literal as possible consistent with giving a sense of the flow of Herodotos’ narrative. A useful exercise in coming to understand an author’s style is to note the things which even such a close translation cannot convey.]

Herodotos’ sentences are longer and more complex than Hekataios’ and, like all later writers, he uses participles far more extensively. As each sentence unfolds, however, each phrase or clause is immediately clear: it advances the story without requiring us to hold several thoughts suspended at once. Thus the first sentence consists of a genitive absolute, followed by the main verb apikneetai, followed by its subject (aner, still unnamed), followed by more about the man: his misfortune, his uncleanliness, his birth. At the end the phrases with men and de, both dependent on eon, lead up to the climactic word basileiou; the chiastic placement of the similar words geneei and geneos accentuates the final phrase.

Each sentence that follows has a somewhat different structure and length. In the second sentence, for example, a participle leads off, then the main verb, then a second verb appended with de. Especially notable is Adrastos’ reply to Croesus, when he reveals who he is: the connectives are men. . .de. . .de. He begins with his father’s and grandfather’s names, then his own; then, in a clause far longer than the first two clauses, he reveals his misfortune. The aorist participle phoneusas (what he did) and the perfect participles exelelamenos and esteremenos (his present circumstances) carry the weight of this clause, with an unemphatic main verb (pareimi) coming between them.

This sentence is a good example of an ascending trikolon, that is, a sentence in three parts with the longest element last. Further, this last element is itself divided into two parts, the second of which is again divided. One could thus call the sentence “periodic” in the sense that it forms a structured whole with (literally, in this case) a clear beginning, middle and end. Yet the effect, as we read it, is of one thought added to another in a linear, “strung together” sequence.

Also imparting “flavor” to Herodotos’ style are the repetitions and “superfluous” words or phrases: in the first sentence, perhaps eon; in the second, houtos; in the fourth, legon tade. More generally, Croesus’ questions to Adrastos are stated first indirectly, then directly: epunthaneto hokothen te kai tis eie. . . .tis te eon kai kothen. (Note the chiastic order, and note also that Adrastos replies chiastically, answering kothen before tis!) The effect is of an unhurried narrative–but not a flaccid one. Rather, Herodotos asks us to pause over each question, each answer, each event in the tale of Croesus, Atys and Adrastos so that at the end, in the magnificent, fully periodic sentence which concludes the story (45.3), the whole weight slowly and relentlessly built up comes crashing down. [Cf. Denniston, Greek Prose Style, p. 8.]

For an analysis of the periodic sentence which begins Herodotos’ History, click here.


Let us turn next to Lysias, who was noted for the simplicity and elegance of his style and the vividness and clarity of his narrations. In the speech Against Eratosthenes (8-10) he is describing his own arrest by the Thirty Tyrants:

For a graphic image of the Greek text, click here.

Dividing up the houses, they went on their way. Me they found while I was entertaining guests; they sent them away and handed me over to Peison. The others went to the workshop and made a list of the slaves. I asked Peison if he wanted to take money and rescue me. He said he would, if it were a lot. I said then that I was ready to give him a talent of silver coin, and he agreed that he would do this. Now I knew that he had no regard for either gods or human beings, but nevertheless in the present circumstances it seemed to me to be imperative to get some pledge from him. After he swore an oath, calling down ruin on himself and his children, that he would accept the talent and save me, I went into the room and opened the chest. Peison noticed and came in, and when he saw the contents he called two of his servants and ordered them to take the things in the chest.

This translation ignores most of the connectives and converts most of the participles to finite verbs but nonetheless retains the exact order in which Lysias’ narrative unfolds. Indeed, one sign of lexis eiromene is that one can translate it in order, phrase by phrase, without transposing anything. Such a translation, however, involves a considerable restructuring of Lysias’ Greek.

To examine further the way Lysias has structured his narrative let us reproduce each element of the Greek on a separate line, with indentations to show the relationship among them; the excruciatingly literal translation which follows will show how complex the “loose” style of Lysias really is. To emphasize the careful connection and subordination of the Greek (even where ideas seem to follow straightforwardly one after the other), words which serve as signposts are in boldface and circumstantial participles, which express so many subordinated ideas so concisely, are underlined in the Greek graphic.

For a graphic image of the Greek text, click here.

And dividing up the houses they went on their way.

And me, on the one hand, they found entertaining guests,

whom driving out they handed me over to Peison;

the rest, on the other hand, going to the workshop, made a list of the slaves.

And I asked Peison, on the one hand, (5)

if he wanted to rescue me, taking money.

he, on the other hand, said he did,

if it were a lot.

I said then that I was prepared to give a talent of silver coin,

and he agreed that he would do this. (10)

Now I knew that he had no regard for either gods or men,

but nonetheless in the present circumstances it seemed to me to be most necessary

to take a pledge from him.

And when he swore,

calling down ruin on himself and his children, (15)

that accepting the talent he would save me,

going into the room I opened the box;

But Peison, noticing, came in

and, seeing the contents, called two of his servants

and ordered them to take the things in the box.

[As a tool for analyzing Greek style such an ultra-literal translation is far more useful than a smooth, fully Englished version, since the goal is to understand the Greek, not to replace it. Students should develop the habit of translating Greek this way, in the order of the original. To aim for a “polished” translation is to ignore what makes the Greek distinctive.]

Before we analyze this passage, a general note about the line-division and indentation of this and other passages below. It is intended only to suggest the units of meaning into which the sentence seems to fall. Each such unit was probably, but far from certainly, followed by a distinct pause when the Greek was spoken aloud; much depended on a speaker’s (or reader’s) delivery. No “scientific” accuracy is claimed or, indeed, is possible. Lines often, but not always, correspond to kola. (To be entirely consistent one would have to place eipon oun [9], for example, on a separate line, but this would create a false impression of composition in very short units [kommata].)

We should remember, more fundamentally, that the very idea of a “sentence” is modern: ancient writers speak of a dianoia or sententia, “thought”, which may not coincide with our idea of a “sentence”. [For a modern definition of a sentence see, e.g., Rickert, New Methods for the Study of Literature, p. 111.] Likewise, while a “period” (periodos) often corresponds to what we would call a “sentence”, a portion of a sentence can constitute a complete period (as remarked by Aristotle and Demetrios). Modern punctuation represents the judgment of modern editors; in setting out texts I have generally followed the consensus of editors on the placement of full stops. [See, e.g., Johnson, Luxuriance and Economy, pp. 15-17 with references.]

Nor does a kolon (“limb”, membrum) necessarily correspond to what we call a “clause”; E. Fraenkel has shown that even a short phrase can have the force of a kolon. [See also H. Dik, Word Order in Ancient Greek, chapter 3.] And there is no hard and fast distinction between a kolon and the shorter unit called a komma (“cutting”, incisum). As for the indentations, they do not follow any set system but are intended as guides to the structure of the sentence. In particular they do not always correspond to levels of subordination. (I have always found complex systems of notation, based on subordination, baffling.) Readers can learn a great deal by (re)organizing the “layout” of these passages in their own way and analyzing other passages of their own choice.

Now for Lysias. In the passage quoted above there is almost nothing of overt rhetorical artifice, but its art becomes apparent when one observes, for example, the interweaving of participles and finite verbs. They tend to be paired, but not mechanically so. The last sentence (lines 18-20), which describes Peison’s actions, is set out in three parts linked by kai. Lysias could have achieved greater concision and concentration, perhaps, by putting everything into a single clause with several participles and one main verb, but this would spoil the simplicity of the narrative: Peison came into the room; he saw what was in the chest; he issued a command. (Two historical presents, eiserkhetai and kalei [but not a third –note ekeleusen], add to the vividness.) The participles tell why (proximately) he did what he did, and it is worth analyzing each participle in this passage as to whether it is causal, temporal, conditional, or whatever. Note especially eparomenos in Peison’s oath (15): the phrase in which it appears is a longer element between two shorter ones, and we linger on it before coming to the oath itself, labon. . .sosein.

Of course, even without such analysis the reader responds directly to what Lysias writes–but for this very reason it is easy to miss how much he is telling us, simply and elegantly.

Notice also the variety. Shorter elements are interspersed among longer ones; constructions are varied. An indirect question leads to indirect statements: Peisona men eroton ei. . . , ho d’ ephasken ei. . . ., eipon oun hoti . . . ., ho d’ homologese. . . ., epistamen men oun hoti. . ., homos d’. . .edokei. . . ., epeide de omosen. . . . Note, too, the placement of sosai. . .labon (6), labein (13), labon. . .sosein (16), labein (20). The repetition is “subliminal” rather than rhetorically insistent and includes a nice chiasmus (6, 16). In short, this is rhetorical art which succeeds by concealing itself. As with Herodotos, lexis eiromene is anything but naive. If one wants to appreciate Lysias fully, one need only attempt to translate a similar narrative from English to Greek!

For an analysis of periodic sentences by Lysias, click here.

Plato’s “Naive” Style

Sophisticated in a different way is the storytelling style which Plato adopts from time to time, for example in the fable about the origin of justice told by Protagoras in the dialogue of that name. Here is an excerpt (320c8-e4):

For a graphic image of the Greek text, click here.

Once there was a time

when gods existed

but mortal kinds did not.

And when for these too there came the appointed time of birth,

gods molded them within the earth (5)

from earth and fire mixing them

and from the things which blend with fire and earth.

And when they were ready to lead them to the light,

they bade Prometheus and Epimetheus

adorn them and assign powers to each as was fitting. (10)

And Epimetheus asked Prometheus if he could assign them himself.

“And when I’ve assigned them,” he said, “you inspect them.”

And so persuading, he assigned.

And in his assigning, to some he attached strength without speed;

the weaker with speed he adorned; (15)

some he weaponed;

to some awarding a weaponless nature,

another power he devised for their safety:

to the ones he wrapped in smallness

winged flight or underground home he assigned; (20)

the ones he increased in greatness,

by this very thing he made them safe.

And so he assigned the rest, maintaining a balance.

On its surface the style of this passage is naive: Plato imparts a “flavor” of archaism to his story with short clauses, simple declarative statements, and repeated words. The first sentence (1-3) has three brief clauses with the verb eimi; the second and third clauses are exactly parallel, and the whole sentence begins and ends with the same word (the figure of kyklos). The variation theoi/thneta gene avoids a too-exact symmetry.

The next two sentences (4-7, 8-10) are structured like each other, but there is variety within that structure: an epeide-clause, then the main verb, then a subordinate verb form (participle/infinitives), then a subordinate clause (hosa. . .hos. . .). [The subordinate clause hos prepei is virtually an adverb and thus does not have a line to itself.] The first of these sentences could have been ended after lines 5 or 6; the addition of line 7 is elegant, with its chiasmus (ges kai puros, 6; puri kai gei, 7) and the variation meixantes/kerannutai.

Three short statements follow (11-13), with a fivefold repetition of forms of nemo carrying over from the preceding sentence (10) and into the next (14). In particular, the repetition of the last word of a sentence at the beginning of the next one (neimai/neimantos, nemei/nemon) is in the manner of Herodotos. Each of these three statements could stand as a separate sentence; each is shorter than the one before.

One can see the art of this by glancing at the next sentence, which runs through line 18 (in thought, all the way to 23): exactly the sort of variety that Plato always seeks, and a sure sign that he is creating the illusion of a naive style, not the real thing. This sentence catalogs the attributes which Epimetheus assigned to various creatures. After the link-word nemon these are set forth with a men and three des; note the chiastic variation tois/tous/tous/tois and the varied length of the clauses. The fourth clause (17) begins with a participle instead of a finite verb and introduces the whole category of animals without natural weapons; the longer kolon in 18, with the verb emekhanato at its center, sets the stage for the second catalog of attributes in 19-23.

Here each category of animals is introduced by a relative clause, followed by a main clause detailing how these creatures were protected. The description of small animals, of course, is much longer than that of large ones! Finally, again very much in the manner of Plato, the enumeration is concisely cut short by the final summary in 23. Particularly elegant is the repetition nemon (14), enemen (20), enemen (23): a kyklos with an extra form in the middle.

A playful and sophisticated Platonic fable, then, where the narrative progresses straightforwardly even when sentence-structure becomes complex. Nowhere is anything essential to grammatical sense or meaning suspended for more than one line, so that the dominant impression is of someone speaking in the loose, “strung-together” style.

Periodic Style: Demosthenes and Isokrates

Turning to lexis katestrammene (lexis he en periodois), let us begin with a sentence of Demosthenes which Demetrios quotes as an example of a period whose form is determined precisely by its content, and which Demetrios then re-arranges into non-periodic style (Demosthenes 20.1; Demetrios 10-11). Here, and in the examples which follow, the periodic effect is gained by suspending until the end of the sentence an element essential for the grammar and the sense.

For a graphic image of the Greek text, click here.

Men of the jury,

most of all because I think it benefits the city for the law to be annulled,

second for the sake of Chabrias’ child,

I have agreed that I shall, as best I am able, speak for these men.

Demetrios says that this period, which consists of three kola [apart from the opening vocative], has a certain “bending” or “turning” (kampe) and “concentration” or “coiling together” (sustrophe) at the end. What gives it this quality? Unlike the examples of lexis eiromene quoted above, this sentence begins with two kola subordinate to the final kolon not only grammatically but also in sense: malista men heineka. . . ./eita kai. . .heineka. . . . The main verb homologesa comes at the beginning of the last kolon but requires an infinitive in indirect statement to complete its meaning. Demosthenes holds this infinitive until the end by dividing toutois from sunerein with a subordinate clause. The whole sentence is aimed at an end, which sunerein provides.

Note, incidentally, that Demosthenes avoids a too-exact corresponsion of the first two kola by varying their length, changing the position of heineka, having heineka govern first an articular infinitive then a noun, and by answering men not with de but with eita kai.

Demetrios’ re-arrangement of this into a non-period is instructive.

For a graphic image of the Greek text, click here.

I shall speak for these men, citizens of Athens.

For Chabrias’ son is a friend of mine,

and much more than him the city,

for which it is right for me to speak.

As Demetrios remarks, the period has disappeared. Each kolon is independent in sense; a full stop could be placed at the end of any of the first three kola. Nothing forces us to look ahead for a conclusion to come. (Indeed, Demosthenes’ final word has become Demetrios’ first.) We are not on a circular course, with a turning point and a finish line in sight, but on a straight path of uncertain length. Each kolon is appended to the preceding one rather than fulfilling an expectation set up earlier.

For excerpts from Demetrios’ essay, click here.

A longer periodic sentence of Demosthenes is analyzed below.

To return to the table of contents, click here.

Here is a period from Isokrates’ Panegyrikos (186), cited by Aristotle in the Rhetoric (1410a):

For a graphic image of the Greek text, click here.

Fame and memory and glory

how much ought one to think

those will either possess while alive

or leave behind when they die

who in deeds of this sort excel?

Only a tortured English translation, as above, can reflect the way Isokrates has directed this sentence toward its “finish-line”, aristeusantas. Each element of the sentence requires something later to complete its meaning: first come three accusatives; then they are made part of a question governed by the verbs khre and nomizein, the first of which requires an accusative subject, the second an infinitive; then come two parallel accusatives and infinitives, but since the accusative words are participles we still don’t know who the sentence is about; finally, the phrase tous. . .aristeusantas, with a prepositional phrase enclosed, tells us.

Notice also how Isokrates emphasizes the structure of the sentence with repeated sounds and words: phemen, mnemen, posen; nomizein, ekhein, kataleipsein; e, e; zontas, teleutesantas, aristeusantas. Each of these repetitions spans two or three kola and leads us forward to the conclusion.

A longer periodic sentence of Isokrates is analyzed below.

Short and Long Kola: Xenophon and Thucydides

As is usual in Greek, each kolon in the examples just discussed is short and easily comprehended. Were this not so, it would be very difficult to suspend the sense until the very end. Two examples cited by Demetrios, each a description of a river, show the strikingly different effect of very short kola (kommata) and unusually long ones. First, Xenophon’s description of the river Teleboas (Anabasis 4.4.3, Demetrios 6):

houtos d’ en kalos men, megas d’ ou.

This (river) was beautiful, but not large.

Then Thucydides’ description of the Acheloos (2.102.2, Demetrios 45):

For a graphic image of the Greek text, click here.

For the river Acheloos,

flowing down from Mount Pindos through Dolopia and the Agraioi and Amphilochoi,

and through the Akarnanian plain,

skirting the city Stratos upland

and discharging into the sea by Oiniadai and creating a marsh around their city,

makes it impossible, because of the water, in winter to mount a campaign.

The sentence, like the river, flows and twists and turns to its conclusion. Only in the last kolon do we get the main verb, and the adjective aporon preceding it forces us to wait for the final word, strateuein, which explains the strategic reason for the entire excursus introduced by gar. An appropriately serpentine periodos!

A Lysianic Period

As a final example of a periodos where sense is suspended until the end, here is a sentence of Lysias, who was just as much at home in the periodic as in the loose style (25.18):

For a graphic image of the Greek text, click here.

But if you think it necessary

the men whom those ones omitted to harm

yourselves to ruin,

none of the citizens will be left.

In the protasis of this conditional sentence Lysias gains extra force first by using ei plus the indicative rather than the more usual ean plus subjunctive (making the whole sentence in effect a future most vivid or minatory conditional sentence), then by bringing the relative clause forward to precede the infinitive apolesai on which it depends. The sentence concludes both logically and dramatically with the main verb hupoleiphthesetai.

A Longer Isokratean Period

Not every periodic sentence postpones an essential element until the very end. Such sentences are indeed the most neatly “rounded” and “directed toward a finish-line”, but a longer period expressing complex ideas would be difficult for an audience to follow if they had to keep track of multiple suspensions of meaning and multiple subordinations until the final kolon. Often the main verb comes at the beginning or in the middle, and the effect of a structured whole is achieved by carefully arranging parallel or antithetical kola whose sense is complete before the sentence moves on. Here is Isokrates’ description of Xerxes’ march through northern Greece (Panegyrikos 89, cited in part by Aristotle, Rhetoric 1410a):

For a graphic image of the Greek text, click here.

Who came to so great a point of arrogance

that, thinking it was a small task to subjugate Greece

and wanting to leave a monument so great

that it is not a thing of human birth,

he did not earlier stop (5)

until he discovered and compelled

the thing which all men talk about:

how with his army

to sail through the mainland

and march through the sea, (10)

yoking the Hellespont

and digging through Athos.

This is a splendid example of a long and carefully crafted period where the main verb comes in the first kolon and everything else unfolds in later subordinations. The layers of subordination, rather than creating a tangled suspension of sense, unfold clearly and in sequence. Note how small the elements are which make up most of this sentence. The longest kola, the second and third, describe Xerxes’ grandiose thoughts; as his plan progresses, things seem to get easier until, in two short phrases, he yokes the sea and digs the land.

And note the careful and varied connections: relative clauses, result clauses, a temporal clause, pairs of participles and a pair of infinitives set off by men and de. Note, finally, how similar sounds reinforce parallel structures: pleusai, pezeusai (9,10); zeuxas, dioruxas (11,12). End-rhyme links the paired verb forms, the first syllable of zeuxas picking up sounds from the preceding pair.

Another period of Isokrates is analyzed above.

More Lysianic Periods

Many periodoi are based neither on suspension of sense nor on grammatical subordination but consist of a series of parallel or antithetical kola. Here are two examples from Lysias cited by C. D. Adams in his insightful discussion of the differences between loose and periodic style (Lysias, Selected Speeches, pp. 345-352):

For a graphic image of the Greek text, click here.

The killing of men they thought of no importance;

the taking of money they considered of great importance. (12.7)

For a graphic image of the Greek text, click here.

These men

many of the citizens to our enemies they banished;

many they wrongly killed and left unburied;

many with civic rights they disenfranchised;

of many the daughters about to be married they balked. (12.21)

Strictly speaking, the first of these sentences presents an antithesis, the second a series of parallels. In each, however, as commonly, the kola are closely parallel in form: in the first, each kolon consists of infinitive + direct object + phrase with peri + main verb; at the end the rhyme hegounto/epoiounto (homoioteleuton) reinforces the parallel and brings closure. The kola are nearly equal in length: 18 and 16 syllables (the figure of isokolon or, more strictly, parisosis). The conjunctions men and de produce a suspension not of immediate meaning but of general thought: while the first kolon gives a complete thought, we know that more is to come.

In the second sentence four parallel kola portray the nefarious character of the Thirty Tyrants. Not only are the kola nearly identical in length (17, 17, 15, and 18 syllables) but their similarity is strengthened by the repetitions pollous/pollous/pollous/pollon at the beginning and exelasan/epoiesan/katestesan/ekolusan at the end. The detail added by the fourth kolon–the confiscations of property even prevented some families from giving their daughters dowries–provides a vivid and dramatic climax and brings a sense of periodic closure.

For excerpts from Demetrios’ essay, click here.

For a graphic of Lysias 12.99-100 laid out in kola and kommata, click here.

A Herodotean Period

Even a single kolon can be periodic. Demetrios cites (17) the opening of Herodotos’ History:

Herodotou Halikarnasseos histories apodexis hede. . . .

Of Herodotos of Halikarnassos’ inquiry the result is this.

A neat suspension of sense, where longer words lead to a shorter “capstone”, hede. Since this period is part of a longer sentence, it is worth quoting the whole thing:

For a graphic image of the Greek text, click here.

Of Herodotos of Halikarnassos’ inquiry the result is this,

that neither may the things done by men with time become extinct

nor may great and wondrous deeds,

some by Greeks,

some by foreigners shown forth,

become obscure,

both the other events and through what cause they warred with each other.

Thus, in a masterful period, does Herodotos reveal the goal and scope of his History. Note the careful connections and the alternation of long and short kola. Note the double purpose clause (hos mete. . .genetai, mete. . .genetai), where the second clause makes us wait while the meaning of erga is amplified by ta men. . . ta de. . . in apposition. Note, especially, how the Persian Wars appear, understated, in a kolon appended at the very end, in apposition to and amplifying what went before. Isokrates could have done no better.

Rhetorical, Historical, and Philosophical Periods

Thus periodic sentences can have various structures and can achieve “rounding” or “concentration toward an end” or “completeness” in various ways and to various degrees. In addition to Adams‘ distinction between periods involving suspension of sense, antithesis, and parallelism one should take note of the more fundamental distinction which Demetrios draws between three types of period, according to how tightly they are constructed: the rhetorical period, the historical period, and the philosophical period or period of dialogue.

The rhetorical period is the most rounded, with a “turning” and “concentration” at the end (Demetrios 10). Examples are the passages from Demosthenes (20.1), Thucydides (2.102.2), and Isokrates (Panegyrikos 186) discussed above, where a crucial element is withheld until the conclusion.

To return to the table of contents, click here.

A Short Historical Period of Xenophon

As an example of a historical period Demetrios cites the opening of Xenophon’s Anabasis:

For a graphic image of the Greek text, click here.

Of Dareios and Parysatis were born two children,

the older Artaxerxes, the younger Kyros.

(Anabasis 1.1; Demetrios 3)

According to Demetrios this sentence consists of two kola, each of which, while an integral part of the sentence, completes its own thought. [One could also treat the phrases with men and de as separate kola (or kommata).] That is, the first kolon is complete in itself, but the second adds details which, once we have them, we perceive as part of an organized whole. Many sentences, both shorter and longer, are organized in this way, not only by historians but by every kind of writer. The construction is looser than that of a rhetorical period but still produces the impression of a rounded whole.

To return to the table of contents, click here.

A Philosophical Period of Plato

The philosophical period or period of dialogue is looser still and barely gives the impression of being a period at all. Demetrios cites the opening of Plato’s Republic:

For a graphic image of the Greek text, click here.

I went down yesterday to Peiraios

with Glaukon, Ariston’s son,

to pray to the goddess

and also wanting to see the festival,

how they would perform it,

since now they’re holding it for the first time.

(Plato, Republic 327a1-3; Demetrios 21)

Demetrios notes that kolon is piled on kolon in apparent artlessness–yet, as critics have long remarked, this is one of the most studied and artful of Greek sentences. [See, for example, Denniston’s remarks (Greek Prose Style, p. 41).] Plato is said to have puzzled long over the order of the opening words. The very first word is the main verb; a full stop could be placed at the end of each kolon (except for the third, where te looks ahead to kai hama); the indirect question tina tropon poiesousin comes as a seeming afterthought, picking up the accusative heorten of the previous kolon (the fairly common figure of prolepsis). Yet by the end we have run a full course, a periodos. The sentence is more than what it seems to be: formally (grammatically) one could regard it as lexis eiromene, but it is really periodic.

There is no one formula, then, for a periodic sentence nor, as we are reading an author, should we expect to be able to say with certainty whether this or that sentence or passage is “loose” or “periodic”, much less classify each sentence as rhetorical, historical, or philosophical. Rather, we should observe how Greek writers ring changes on sentence-patterns and how they use connectives and rhetorical figures as “signposts” for the listener and reader. Labels, where we can assign them, are means, not ends.

For Demetrios’ remarks on the three types of period, click here.

A Period of Gorgias

In order to place in perspective the use of rhetorical figures as guides to the architecture of a sentence, let us look at the most exuberant rhetorician of them all, Gorgias of Leontinoi, who amazed the Athenians with a style in which repeated sounds and plays on words were part of the very texture of the Greek. Our Gorgianic exemplar is chapter 6 of the Encomium of Helen.

For a graphic image of the Greek text, click here.

For either by Fortune’s volitions

and the gods’ counsels

and Necessity’s decrees

she did what she did,

or by force seized (5)

or by words persuaded

{or by love captured}.

If, then, because of the first,

worthy to be accused is the one accusing:

for god’s zealousness by human forethought it is impossible to hinder. (10)

For it is natural not for the stronger thing by the weaker to be hindered,

but for the weaker by the stronger to be ruled and guided,

and for the stronger to lead

and the weaker to follow.

But god than man is a stronger thing (15)

both in force and in wisdom and in all else.

If, then, to Fortune and the god the blame one must assign,

verily Helen from her ill fame one must absolve.

Not only is the passage composed for the most part of very short elements (kommata rather than kola) but many of these are the same or nearly the same in length (isokolon or parisosis): Tukhes boulemasi, theon bouleumasi, and Anagkes psephismasin (lines 1-3) are 6, 6, and 7 syllables; to men kreisson hegeisthai and to de hesson hepesthai (13-14) are each 7 syllables.

End-rhyme (homoioteleuton) is everywhere: boulemasi, bouleumasi, psephismasin; harpastheisa, peistheisa, (5-7) [{e eroti halousa} is a medieval conjecture based on Gorgias’ development of the argument in chapter 15]; anatheteon, apoluteon (17-18).

Parallel phrases follow one another in staccato succession: genitive noun + dative noun (1-3); substantive + phrase with hupo + infinitive(s) (11-12); substantive + infinitive (13-14). Parallelism of sound and structure comes from plays on sound (parechesis) and meaning (paronomasia) within and between phrases, with similar words and even forms of the same word juggled kaleidoscopically: boulemasi/bouleumasi (1-2); aitiasthai/aitiomenos (9), where the first form is passive, the second middle; kreisson. . .hessonos. . .hesson. . .kreissonos. . .kreisson. . .hesson (11-14), a double chiasmus. Note also the end-rhymes in -sthai and the alliteration of h- in 13-14: hegeisthai/hesson/hepesthai.

Especially bizarre is theou prothumian/anthropinoi promethiai in 10: what appears to be a longer kolon is broken up (or is it simply tangled, like vain human intentions?) by these nearly impenetrable phrases.

Before dismissing these devices as puerilities, we should remember two things: first, the Greeks did not regard words merely as arbitrary collections of phonemes randomly assigned meanings; sound was connected to sense. Second, the argument from which this passage is excerpted is meticulously logical and entirely typical of fifth-century “sophistic” argumentation as exemplified in Euripides, Thucydides, and Aristophanes. [See the works of John Finley cited in the bibliography.]

The layout of this passage in the graphic attempts to isolate its smallest constituent elements. Gorgias encourages this with his insistent repetitions and isokola. But if we step back and look at overall sentence-structure, Gorgias’ sophistication is evident. The first sentence (1-7) pivots around epraxen ha epraxen; the possible reasons for Helen’s elopement are presented on either side of this main clause. Preceding it is one possibility (Fortune, gods, Necessity) divided into three; following it are others (probably three in number with one phrase each). Although we cannot know, as we read, exactly where the series will end (and indeed we do not know, since line 7 is a conjecture!), Gorgias suspends the sense nicely near the center of the sentence.

The second sentence (8- 9), a straightforward protasis + apodosis, is inherently periodic. In the next (10) the predicate adjective adunaton, which reveals the grammatical construction, comes next to last, and the direct object prothumian receives its infinitive (the subject of the sentence) only at the end.

By contrast, in the longer sentence which follows (11- 14) everything depends on the first word, pephuke: the first kolon, introduced by ou, leads to the second, which answers ou with alla and where koluesthai is answered by two infinitives, arkhesthai and agesthai; at the end (13-14) two further articular infinitives amplify the idea expressed in line 12. Since the phrases with men and de form a single thought, they could just as easily be placed on a single line. They would then complete an ascending trikolon.

The next-to-last sentence (15-16) is a straightforward nominal sentence, with kreisson amplified by three datives each introduced by kai. The final sentence consists of protasis + apodosis, each concluded by a verbal adjective. [In line 18 the first word is MacDowell’s conjecture (CQ [n.s.] 11 [1961] p. 121).]

Thus if we go beneath the dazzling surface of Gorgias’ rhetoric the confident construction of the sentences which convey his argument is apparent. Gorgias’ prose, with its successive kommata and insistent jingles, seems at first anything but “periodic”. Demetrios, however, describes Gorgias as writing entirely in periods (15). Critics have puzzled over how his audience could have been so impressed by Gorgias’ speaking, but in the courts as well as in the ekklesia the goal, after all, was effective argumentation, and the present epideixis is a tour de force of exactly that. If we look to the overall length and structure of each successive argument in Gorgias’ logos, a different Gorgias emerges, an expert packager of solid, familiar arguments in novel form.

Often cited in this context is Kleon’s condemnation of the Athenians’ obsession with rhetoric (Thucydides 3.38). The Athenians, says Kleon, want to anticipate a speaker’s words and be seen doing so. Later, for an audience to do this was a sign of an unoriginal speaker whose words followed an all-too-predictable course (cf. Demetrios 15). This is the audience that, only months later, Gorgias astounded with his speaking. Such a seasoned, even jaded audience was ready for something new. Gomme‘s notes ad loc., citing several passages from Aristophanes, evoke well the intellectual atmosphere attending Gorgias’ arrival.

To conclude this section here is an appropriately rhetorical translation of this passage by Larue Van Hook:

For either by the disposition of fortune and the ratification of the gods and the determination of necessity she did what she did, or by violence confounded, or by persuasion dumbfounded, or to Love surrendered. If, however, it was against her will, the culpable should not be exculpated. For it is impossible to forestall divine disposals by human proposals. It is a law of nature that the stronger is not subordinated to the weaker, but the weaker is subjugated and dominated by the stronger; the stronger is the leader, while the weaker is the entreater. Divinity surpasses humanity in might, in sight, and in all else. Therefore, if on fortune and the deity we must visit condemnation, the infamy of Helen should find no confirmation.

[Published in his introduction to Isokrates’ Helen, in vol. 3 of the Loeb edition of Isokrates, pp. 55-57.]


A Final Demosthenic Example

Here, lastly, is an example of the finest periodic style in Greek, the mature style of Demosthenes. Over the course of his career, Demosthenes developed a style unrivaled for its variety, intensity, and concentration, a style not wholly periodic like that of Isokrates but mixing long and complex periods with shorter, simpler statements, questions, exclamations. A good Demosthenic periodos, often cited, is the opening of the speech On the Crown:

For a graphic image of the Greek text, click here.

First, Athenian men, I pray to all the gods and goddesses,

as much good will as I continually have to the city and all of you,

that so much be accorded me from you for this contest;

second, a thing which especially concerns you and your piety and reputation,

this for the gods to grant you: (5)

not to make my opponent your advisor

about how you ought to listen to me

(that would be outrageous)

but the laws and the oath,

in which, in addition to the other just things, this too is written: (10)

to listen to both in the same way.

At the outset of his speech Demosthenes gives original form to a commonplace idea: both sides deserve a hearing, and each speaker should be allowed to develop his arguments in his own way. At each stage of the sentence grammatical “signposts” and other key words create a suspension of meaning which is only resolved in a later kolon. Demosthenes casts the entire sentence as a prayer (as he does the conclusion of the speech [324], which repays comparison).

The prayer has two main parts, introduced by proton men (1) and epeith’ (4). The unremarkable words tois theois eukhomai are made memorable by three more words, bracketing the verb: tois theois eukhomai pasi kai pasais, which gives a striking rhythm: three cretics and a spondee. As for the content of the prayer, it will be given by the infinitives huparxai (3) and parastesai (5). Preceding the first is a relative clause introduced by hosen, which looks ahead to the correlative tosauten for the completion of its meaning: in return for his efforts on behalf of the city and its citizens, Demosthenes asks an equal recompense (a point reinforced by humin. . .par’ humon). The second infinitive, parastesai, is again preceded by a relative clause, this time introduced by hoper, which looks ahead to its antecedent touto (and note, in 4 and 5, huper humon and humin, forming an elegant chiasmus with the earlier pronouns humin [2] and humon [3]).

But the second part of the prayer is not complete. Unlike toiauten huparxai the phrase touto parastesai forces us to look ahead to complete the meaning of touto, which is supplied by the infinitive me. . .poiesasthai, in apposition to touto. But we are not through yet. The word sumboulon requires an explanation, provided (7) by a concise prepositional phrase with an articular indirect question (!). Demosthenes then pauses for a parenthetical remark (8). Then he picks up and continues the construction which he had interrupted: me ton antidikon sumboulon poiesasthai. . .alla tous nomous kai ton horkon. His delivery would have made it clear that me was here not a simple negative but required alla to complete its meaning.

The period concludes with a relative clause explaining the jurors’ oath–a clause whose culmination is prepared for by a prepositional phrase with pros and by the pronoun touto, in apposition to which comes the final articular infinitive–the point of the whole sentence, an equal hearing for both sides, invested now with the authority of the gods and the solemnity of the oath.

A complex period, with many suspensions of sense. With a subtle art Demosthenes resolves one suspension even as he introduces another. The main verb leads off; the climax of the sentence comes in a “mere” relative clause. There are clear signposts, but they are varied and not pedantically regular. The same variety is apparent in the two parts of the prayer (2-3, 5-11). Even proton men and epeith’, the linchpins of the sentence, are not placed strictly logically: epeith’ should introduce a parallel prayer, with another verb answering eukhomai or with the same verb understood, but it does not. It introduces the second prayer dependent on eukhomai. Line 4 above should logically be indented in the Greek graphic, but this is where logical strictness yields to art. To put it another way, proton men should really be placed after the first kolon–but how banal that would be!

A shorter period of Demosthenes is analyzed here.

More Periods of Gorgias, Isokrates, and Demosthenes

Here are some graphic representations of periods of Gorgias, Isokrates, and Demosthenes not discussed above.

Click here to see Gorgianic periods.

Click here to see Isokratean periods.

Click here to see Demosthenic periods.

Isokrates, Demosthenes, and Cicero

The distinction between the periodic structures favored by Isokrates and those of Demosthenes is brought out well by Eric Laughton in his comparison of these orators with Cicero, who is Demosthenic rather than Isokratean (“Cicero and the Greek Orators”, AJP 82 [1961] 27-49; see especially pp. 41-49). Cf. H. C. Gotoff, “Thrasymachus of Chalcedon and Ciceronian Style.”

But when one looks back at the elegant repetitions and subordinations, at the effortless variety of Plato’s expression, and especially at the elaborate edifice of 14-23 (nemon . . . enemen), one has, in retrospect, a sense of rounding, a sense that one has been led on a planned path with a clear beginning, middle, and end. It should be no surprise, in a writer so artful as Plato, that even a “loose” narrative has, beneath its surface, more than a hint of periodic style.

[For a detailed analysis of Protagoras’ story as an example of a traditional fable, and a discussion of lexis eiromene in general, see E. Norden, Agnostos Theos, Anhang VII, pp. 366-379. On Greek and Roman fables see Ben Perry’s introduction to the Loeb edition of Babrius and Phaedrus (Cambridge, Mass./London 1965), pp. xi-xxxiv.]