Appendix: The Formality Hypothesis
A tradition stretching back to antiquity categorizes the earliest preserved Attic oratory as “dignified,” “formal,” or “austere.”  In The Evolution of Greek Prose Style Dover offers the most precise analysis of this feature and seeks to explain it as a continuation of the traditional style of public speaking at Athens, the only area that has provided specimens suitable for close study. That style avoided various colloquialisms, including features of oral narrative. Most of Dover’s evidence is persuasive, but in his treatment of oratory he turns his attention to professional speech exclusively. He does not speculate on (or even mention) amateur speech and does not offer a specific motive for the logographoi to retain the basic character of traditional public speech as he reconstructs it from Thucydides, tragedy, and a few other texts.
In arguing for his “hypothesis that the language of early oratory was recognizably formal” Dover adduces Aeschines 1.25, which I quote more fully than he does:Aeschines fatuously stakes his claim on a supposedly archaic statue of Solon,  and the other texts he cites (the more or less identical Aristotle Constitution of the Athenians 28.3 and Plutarch Life of Nicias 8.6) are, as Dover acknowledges, “unsatisfactory” evidence for ancient decorum; but he does regard them as “good evidence for the existence of an ideal by which the actual could be (polemically) judged” (1997:66). As I see it, even the myth of a dignified speech style does not tell us much about the history of the genos dikanikon. All the figures supposedly paradigmatic of a restrained style are leading politicians whose best-known speeches would have been largely or exclusively political, not dicanic, and Aeschines refers to them as rhêtores, a term that would not apply to the obscure Athenians who found themselves needing to speak in court. 
καὶ οὕτως ἦσαν σώφρονες οἱ ἀρχαῖοι ἐκεῖνοι ῥήτορες, ὁ Περικλῆς καὶ ὁ Θεμιστοκλῆς καὶ ὁ Ἀριστείδης, ὁ τὴν ἀνόμοιον ἔχων ἐπωνυμίαν Τιμάρχῳ τουτῳί, ὥστε ὃ νυνὶ πάντες ἐν ἔθει πράττομεν, τὸ τὴν χεῖρα ἔξω ἔχοντες λέγειν, τότε τοῦτο θρασύ τι ἐδόκει εἶναι καὶ εὐλαβοῦντο αὐτὸ πράττειν.
And those public speakers of old, Pericles and Themistocles and Aristides (who bore a title quite unlike that of this man Timarchus – he was known as “the just”) were so decent that in their day this habit that we all practice nowadays, of speaking with the hand outside the clothing, was considered something brash, and they avoided doing it.
I endorse Dover’s insistence that “[f]or the historian, the fact that data of the highest importance [exhibiting the style of symbouleutic and forensic speech prior to about 425] are irrecoverably hidden in darkness is extremely unsatisfactory; but we must never allow that patch of darkness to slip out of our field of vision, never treat what cannot be investigated as if for that reason it did not matter” (1997:60). But it seems to me that, this methodological manifesto notwithstanding, he has not exploited the evidence for the speech of idiôtai and the precise motive those men’s quotidian speech style gave the logographoi for the development and use of a less emotional style.
In one particular I think Dover (1997:63–64) has misinterpreted the data in such a way as to exaggerate the degree to which early professional speech writers reproduced the formality of the tradition: the use or avoidance of the long accented iota suffix attached to demonstrative pronouns and several adverbs, usually (at least in the U.S.) referred to as the “deictic iota,” but by Dover as the “demonstrative affix -ι” There are no such deictic iotas in Athenian public speech as represented in Thucydides and no credible examples in tragedy (Bers 1997:140 with n36). As Dover shows in his statistical table, the ratio of word tokens to which the deictic iota is added to the same tokens without the suffix is highest in Aristophanes (1:4), lowest in early Platonic dialogues (1:30), quite low in Lysias (1:17), Isocrates (1:28), and Isaeus (1:16), but quite high in Demosthenes (1:7.5). There are several reasons to dispute Dover’s conclusions: (1) We can scarcely call the deictic iota rare in the earliest oratory. Antiphon shows the same ratio as Aristophanes, although he is the logographos whose language most often among the orators retains conservative features (still using Ionic forms like -σσ, anchoring μέν and δέ with τοῦτο, etc.),  preferred οὑτωσί to the word without -ί even more than Aristophanes. Lysias uses νυνί more often in proportion to νῦν than Aristophanes. ὡδί makes 14 appearances in Aristophanes and none whatsoever in the canonical orators (a slightly wider group than that canvassed by Dover) other than Demosthenes (12 occurrences), but Andocides is the only orator for whom the TLG shows ὧδε, and in him only once (at 1.25), so it is unlikely that abstention from ὡδί before Demosthenes has any significance. (2) The high frequency of -ί in Aristophanes is very likely to reflect the comic poet’s intention to keep his dialogue highly animated. 
[ back ] 1. The doxography being quite well known, I will mention here only one ancient and one modern (nineteenth-century) opinion linked to it, exemplorum gratia. Dionysius of Halicarnassus, in On Literary Composition (22.29), chooses Antiphon to represent oratory in a short list of authors who best represent the austere style. Jebb 1876:25, who refers to Dionysius, speaks of “the superb decorum of the old school.” To his credit, Jebb leans toward rejection of a moralizing interpretation of the changes that some came along (“It was only when fierce passion and dishonesty had become strong traits of a degenerate national character that vehemence and trickiness came into oratory”). Jebb proposes instead, “It appears simpler to suppose that the conventional stateliness of the old eloquence altogether precluded such vivacity as marked the later; and that the mainspring of the new vivacity was merely the natural impulse, set free from the restraints of the older style, to give arguments their most spirited and effective form” (1876:29).
[ back ] 2. A piece of pseudo-evidence deflated by Demosthenes (19.251–252). The statue, he says, was created so long after Solon’s time that not even the sculptor’s grandfather could have laid eyes on Solon.
[ back ] 3. Note also that in this very encomium of a restrained style Aeschines uses the deictic iota in the word νυνί, a feature that Dover regards as colloquial: see the discussion below.
[ back ] 4. I note a small error in Dover’s table: there are three instances of νυνί in Antiphon (5.90; 6.29, 42 – all in actual speeches).
[ back ] 5. The use of deictic iota is not an Aristophanic tic, but a feature he shares with other comic poets of the fifth century and beyond.