We must always be seeking better rituals and conventions; but we moderns tend to gag at these as stifling to freedom. So we Romantic moderns, especially we Americans, tend to see the issue as simple: rituals and conventions are likely bad, as they are almost certainly not good. Cowboys like things to be so simple. At some level of our common cultural judgments, inherited from our dual heritage stemming from both Puritanism and the Enlightenment, we see ritual and convention as oppressive Catholic priestcraft, or else as either Monarchical or Aristocratic elitist oppression. It is simply in the water here — even if one were to be an American Catholic Monarchist.
Plato was also quite wary of rituals and conventions of a sort that he called “poetry” (ποίησις, from ποιέω “to make, to show, to put/place”), although he practiced a form of it. Although his concerns about convention and “poetry” come from a different place than our concerns about convention and ritual, there are important lessons for us both where his concerns overlap with ours and where they do not overlap.
We are not ourselves terribly troubled by what we call poetry — which we see as perhaps an expensive or eccentric taste at worst, and as a liberating possibility for the human spirit at best. For us, it is decidedly not conventional, or ritual. For Plato, however, Poetry was something very different; when we translate the word ποίησις as “poetry” we collude with an infelicitous conflation of two very different enterprises. We consider “poetry” as part of the “arts”, but the ancient Greeks didn’t have a word to designate the group of disciplines and activities that we would call “art”. Instead, they had the word “τέχνη” (“skill”), which would cover the range of “τέχναι” from medicine to ship-building to masonry to cooking to farming to dancing to making love to poetry, &c.  Disambiguation is helpful. Were we to first trace some of the historical backdrop that occasioned Plato’s concern, we might be in a better place to understand Plato’s Socrates, only then later to see an overwhelming number of analogues in our own world.
Before we look at Plato, then, let us sketch a few outlines of the nature of “poetry” prior to him.
It may seem amusing to sketch outlines of poetry before Plato by starting with Friedrich Nietzsche (A.D. 1844-1900), but we shall be helped by such a starting point. Nietzsche saw the origins of poetry in a technique for mastery, as a skill having utility for nearly every activity, even control over fate. The rhythm of a speech aided in memory, and had an elemental power of coercion in it:
not only the stride of the feet but also the soul itself gives in to the beat [of poetry’s rhythm] — probably also, one inferred, the souls of the gods! 
It could also, as with the Pythagoreans, drive out passions and harmful daemons, bringing the soul back into equilibrium, calming the targeted god (who had wrought harm by his destabilizing frenzy) by over-exciting him, and thereby pacifying him.
For farmers and fishermen, it was another tool: the
mundane song of the most ancient times [harbors a] presumption that rhythmic quality exercises a magical force; when bailing water, for instance, or rowing, the song is a bewitchment of the demons believed to be at work here; it makes them compliant, unfree, and a tool of humans. And whenever one acts, one has an occasion to sing — every action is tied to the assistance of spirits: incantation and conjuration seem to be the primordial form of poetry. 
Indeed, if “all things are full of gods” (Thales), then one cannot risk being without this incantatory power. So too with oracles — the precise recitation of the poetic form of speech was thought to coerce the future away from the Fates via Apollo.
In short: was there anything more useful than rhythm to the old superstitious type of human being? One could do everything with it: promote some work magically; compel a god to appear, to be near, to listen; mould the future according to one’s own will; discharge some excess (of fear, of mania, of pity, of vengefulness) from one’s soul, and not only one’s own soul but also that of the most evil demon. Without verse one was nothing; through verse one almost became a god. 
This is the basic sensibility one finds in the prophecies delivered by oracles. If the reader is struggling to see a stronger connection between all of this and Homer than the enchanted world they hold in common, I would remind him or her that the hexameter (which Homer uses) was supposed to originate at Delphi, where oracles were delivered.
It is not clear that modern scholars would look terribly kindly upon Nietzsche’s historical narratives here. (Although, seemingly in support of it: I do remember reading, in Bradley Birzer’s delightful book on Tolkien, about a reported conversation in which J.R.R. himself mentioned to an interlocutor that the pagan Germanic tribes would say certain spells when they would pick certain flowers, plants, and herbs, to ward off the gods and goddesses associated with them, but that after Christianization, these spells were replaced with Pater Nosters, &c. The textual support for this is in footnote [4a].) There may, perhaps, be little evidence for these stories, or little merit in interpreting the historical evidence this way. Thou, O Reader needest not take this as anything more than a dream of history — a dream that is perhaps at odds with what we shall see in Havelock in Section II., but also certainly entirely similar to Havelock’s portrait with respect to the concerns of Plato’s Socrates.
Here, in the kind of “poetry” Nietzsche recounts, anyone can learn the proper formula; anyone might think to gain access to this kind of power. This is not about civilly soliciting, or dialoguing, or discoursing with elements, spirits or the Fates: rhythm and rhyme are here techniques for coercion. The act of understanding only applies to the execution of the technique. As it is merely a kind of technology, it is not unlike Sophistry; and dovetails with an ultimately irrational view of the world that is quite at odds with the one found in Plato.
Perhaps we reject everything that Nietzsche wrote above — perhaps we would say that the cultural value placed on poetry, rhythm, and rhyme, does not originate in their perceived utility, in functional attempts to gain power or see one’s will accomplished. Certainly Nietzsche acknowledges that the rhythmic nature of our biology inclines us to liking and using it; he also claims, however, that we project this feeling onto the gods, and onto the spirits of the elements, and attempt to use our sense of compulsion and captivity to move them. Perhaps we’d like to resist this biological foundation, and any utilitarian beginnings for poetry; perhaps we’d like to say that our attraction to rhythm gains momentum and finds itself practiced because of the aesthetic encounter or somesuch. Were we to so reject this portrait of utility and power, however, we would need to have good grounds for it, and to be familiar with ancient history. Most people who would object to this would do so simply because they find this description distasteful.
In his book, Preface to Plato, Eric Havelock suggests that poetry, which we experience as an “esoteric” state of mind —i.e., removed from the functional day-to-day of the workplace or the operations of domestic life or the procedures of administration– is in fact, for the pre-literate “Homeric State of Mind”, not esoteric at all, but everyday, and pragmatic.  Not too far from Nietzsche’s portrait, poetry is here functional, here about effective power. Laws were shaped “in the ear”, were “framed orally for verbal memorization” , and when laws or correspondence were written down, the rhythms, the ring forms, &c. were all in the service of the aural functions of preserving speech. Mandates and correspondences were “composed for recital, rather than for reading.”  The newer technology of scrolls preserved speech just fine; and yet the habits of a pre-literate culture persevere even after the introduction of the technology of writing. The poetic and prosaic are not at odds in Homer’s world: there were no prose originals. Poetry would be remembered; prose would not be remembered. Likewise, to be preserved, mandates need to be framed according to the laws of memory, mnemosyne (μνημοσύνη).
In short, all significant communication without exception was framed to obey the psychological laws of the goddess Mnemosune. 
Indeed, these laws, in a pre-literate culture, are not ornamental, but are functional, and efficiently so. He recounts several reports of British soldiers who had fought against pre-literate Turks at the beginning of the twentieth century. During one ceasefire, the Brits greeted the Turks with short phrases (“good luck, chap”), but the Turks greeted the British soldiers with rhythmic proverbs (“smiling may you go, and smiling may you come again”); during one fight, the Turks had an imam shouting improvised verses behind them, to organize them. In a pre-literate culture, this is highly efficient. “The epic style was therefore a necessity for government, and not just a means of recreation.” 
There are two interrelated consequences of this kind of culture, according to Havelock.
The first consequence concerns the nature of education. I quote him at length:
In such non-literate cultures the task of education could be described as putting the whole community into a formulaic state of mind. The instrument for doing this was to use the tribal epics as a paradigm. Their style is intensified to be sure. Their idiom shows a virtuosity which in common transactions might be imitated but at a simpler level of artistry. A minstrel would be a man of superior memory, and so also might be the prince and the judge. This automatically meant superior rhythmic sense, since rhythm was the preservative of speech. With superior memory and rhythmic sense would go also a greater virtuosity in the management of the formulas. The lesser memories of the populace would be content to use simpler and less elaborate language. But the whole community from minstrel and prince down to the peasant was attuned to the psychology of remembrance. […] the epic poet […] would exercise a degree of cultural control over his community which is scarcely imaginable under modern literate conditions in which poetry is no longer part of the day’s work. 
The second consequence results from a poet holding sway over speech via education: such a one also holds sway over thought. Mimesis, imitation, is not only how the preservative of rhyme carries forward: it is also how people learn. The content of what is memorized and imitated has serious consequences for the kind of culture that emerges from this content, has its anchor it it. Another quote, mercifully shorter:
Control over the style of a people’s speech, however indirect, means control also over their thought. The two technologies of preserved communication known to man, namely the poetised style with its acoustic apparatus and the visual prosaic style with its visual and material apparatus, each within their respective domains control also the content of what is communicable. 
Homer was not selected by cultural engineers to be the material with which a population was to be indoctrinated: it is simply in the nature of rhythm and poetry and a particular phrase’s endurance over time to hold sway. Such a thing as rhyme and poetry cannot help but be the training and education of a whole people. Although he did not write for the purpose of educating, Homer himself seems to be aware that this is one of the functions of poetry, as he does, on occasion, include advice about the durability of different kinds of wood, and supplies the definition for arcane words; also, sometimes, “the glorification of the past has an educational by-purpose”. 
This by-purpose is stronger than Homer seemed to have recognized, however. Rhythmic speech is, here in Havelock’s portrait, again a technique, and a technology. It is about securing outcomes, and perpetuating institutions. It is not about understanding.
Before we might profitably begin to understand Plato’s concerns about “poetry”, we should first look at poets and statements about poets in the writings attributed to Homer.
Homer’s Iliad (ca. 9th-7th century B.C.) opens with an invocation of the goddess, and the poem itself is, ostensibly, the singing of the goddess herself as much as a technical feat flowing on the lips of the poet. Homer never appears as a character in The Iliad, though there are rare occasions when his voice –the narrator’s voice– breaks through, even if only as a vocative address to a character. Homer does not appear in The Odyssey either, however, we do see a poet at his craft in that work — Demodocus, performing in the palace of Alcinous, to entertain the anonymous Odysseus. 
[…] summon the godlike singer of tales,
Demodocus. For the god has given him,
Beyond all others, song that delights
However his heart urges him to sing.
This world is full of gods. For the divine singer (θεῖον ἀοιδὸν), the urging of his “heart” (the preferred translation for θυμὸς) had been given (δῶκεν) by the god so as to please (τέρπειν). Yet the gods here are not simply beneficent in the manner that Plato requires in Book Two of The Republic. Instead, Homer has it that
The Muse loved this man [Demodocus], but gave him
Good and evil both, snuffing out the light
Of his eyes as she opened his heart to sweet song.
It is possible that this blindness is part of the blessing –not being able to see, the poet may, perhaps, have a better memory– and the faculty of memory is key for a bard, who must virtuosically weave novel sentences from a hoard of standard phrases. Regardless, once escorted into the halls of Alcinous, Demodocus sits “on a silver-studded chair / propped against a column” with the lyre he shall use suspended from a peg. “The Muse moved the bard to sing of heroes”, and so Demodocus, wine goblet by his side, sings stories recounted in The Iliad, and repeatedly brings Odysseus to tears.
Demodocus paints a picture, and the vividness of this picture is what commends him to Odysseus. The bard is praised by Odysseus because it is ‘as if he were there [him]self, or had heard it from another’: “ὥς τέ που ἢ αὐτὸς παρεὼν ἢ ἄλλου ἀκούσας” [VIII.491]. Odysseus asks the bard to sing about when the Greeks deceived the Trojans into taking the Trojan horse into the walled city of Troy. So the bard, moved or literally “rushed” by the god (“ὁρμηθεὶς”, from “ὁρμάω”), “showed his song” (“φαῖνε δ’ ἀοιδήν”).
The Muses are the daughters of Memory. There has been contention as to whether the bard is participating, in some ecstatic fashion, in the Muse’s direct apprehension of the event (Odysseus grants that it is “as if [the bard] were there [him]self”), and is thus “showing” this direct apprehension in his song, by which the audience is “shown” the events, or whether the god-fired heart of the bard is simply given this Muse-ical ability by nature, and cultivates it by technique and skill — which he could do should he have “heard it from another”, which is the other alternative that Odysseus puts forward. That the bard seems to have power to craft his words –that is, that he is not in a state of mindless frenzy or selfless ecstasy– also seems clear, given that Demodocus not only received the gift from the goddess, but that he sings “whatever his heart rouses him to sing” about (“ὅππῃ θυμὸς ἐποτρύνῃσιν ἀείδειν”).
There are thirty-six epithets for Achilles […], and the choice is rigorously determined by the position in the line and the required syntactical form. It has been calculated that there are some twenty-five formulaic expressions, or fragments of formulas, in the first twenty-five lines of the Iliad alone. About one-third of the entire poem consists of lines or blocks of lines which occur more than once in the work, and the same is true of the Odyssey. […] [T]he repeated formula is indispensable in heroic poetry. The bard composes directly before his audience; he does not recite memorized lines. […] Out of these building blocks [of stock phrases for myths and familiar events of ordinary life] the poet constructs his work, and each work –each performance, in other words– is a new one, though all the elements may be old and well known. 
The phrases that appear become rhythmic: when someone dies, how often does he do so “clutching the earth, which nourishes us all”? And so on. This is why the bard Phemius who we see later in The Odyssey says, in XXII.340ff., “I am self-taught, and a god has planted in my heart / All sorts of songs and stories”  One scholar notes:
This sounds contradictory, but it is difficult to imagine different objects of the teaching and the inspiration. […] [T]he term ‘self-taught’ refers to the singer’s ability to use a traditional technique in a personal way […], but this freedom is not absolute: it is accompanied by a corresponding guidance by the Muse. […] [T]here is no clear delimitation of roles. Hesiod first tells us that the Muses ordered him to celebrate the gods, and then asks them to celebrate the gods. 
Inspiration and technique do not seem to be at odds in the pre-Platonic Greek tradition.  This doesn’t mean that poetry is a conversation, or is open and collaborative. The poet floods the audience with images — and, seemingly, the goddess, the daughter of Memory, also so floods the poet. Being flooded with images does not lead to the audience understanding: the audience is merely disposed in a certain way through the images.
There is something about the power of the Muse in the god-fired bard that attracts, captivates, enchants, has loveliness or charm (χάρις): it is not mimetic fidelity that makes the bard so, for the bard is not writing on a parchment, but performing: his words, drawn from memory, are live, living: his stage presence is key. So Eumaeus, the swineherd, to Penelope (Odysseus’ wife), concerning the mysterious guest who has just arrived (really Odysseus, disguised):
It was just as when men gaze at a bard
Who sings to them songs learned from the gods,
Bittersweet songs, and they could listen forever —
that’s how he charmed me when he sat in my house.
There is a strong fusion, here, between what is compellingly beautiful and what is true. “In this sphere of thought there is no clear distinction between external appearance and intrinsic value.”  The possibility that something charming and powerful might be deceptive hardly appears in Homer, and some would champion that it does not appear there at all; and yet, awareness of the possibility grows as the centuries move on towards Plato. “It was Hesiod who raised the problem of poetic illusion”.  The concern crescendos in Pindar.
Finally, because these words come from the goddess who sees truly, despite the involvement of the poet’s craft in the choices of phrasing, their truth is guaranteed, has authority; it has authority, and this is tied up with its beauty, its traditionality (rhythm is the preservative of speech, and what is preserved is repeated, has more history than a lifetime), and the “word hoard” of stock phrases the bard draws upon, in which the audience has been formed, according to which the culture thinks and judges, and to which it is tuned.
The Homeric bard takes requests, but as an image-painter, as god-fired, there is no substantial interaction with him. Indeed, there is a charming power in him to captivate his audience. In the end, one is to suffer him, one is to be receptive to and moved by him. He may take requests, but he is not moved by questions, but by the story, the performance, the act of captivating for the sake of the narrative. He does not engage in a dialogue with anyone in the audience. It is not merely that the relationship is asymmetrical: it is as one-sided as a jukebox, or perhaps a movie on DVD/Blu-Ray. He performs, and the audience is passive.
Plato on Poetry, ed. Penelope Murray (New York: Cambridge, 1997), 1.
Nietzsche, The Gay Science, transl. Josefine Nauckhoff, ed. Bernard Williams (New York: Cambridge, 2001), 84, Book Two §84
The Gay Science, 85, §84
The Gay Science, 85, §84
So in “Selections From the Corrector and Physician of Burchard of Worms” (ca. A.D. 1008-1112) in Medieval Handbooks of Penance: A Translation of the Principal Libri Poenitentiales and Selections from Related Documents, ed. John T. McNeill & Helena M. Gamer (New York: Colombia, 1990), 330-331, we read
65. Hast thou collected medicinal herbs with evil incantations, not with the creed and the Lord’s prayer, that is, with the singing of the “credo in Deum” and the paternoster? If thou hast done it otherwise  [than with the Christian formulae mentioned] thou shalt do penance for ten days on bread and water.
confirming Tolkien’s story.
Eric A. Havelock, Preface to Plato (Cambridge, MA: Harvard Belknap, 1963), 134
Havelock, Preface, 136
Havelock, Preface, 137
Havelock, Preface, 138
Havelock, Preface, 139
Havelock, Preface, 140
Havelock, Preface, 142
W.J. Verdenius, “The Principles of Greek Literary Criticism”, Mnemosyne, Vol. XXXVI (1983), 32-34
Homer, The Odyssey transl. Stanley Lombardo (Indianapolis: Hackett, 2000), 106ff.
M.I. Finley, The World of Odysseus, introd. Bernard Knox (New York: New York Review of Books, 1982), 22, 23
The Odyssey, transl. Lombardo, 347
See Penelope Murray, “Poetic Inspiration in Early Greece” J.H.S. 101 (1981), 87-100.
The Odyssey XVII.518-521, transl. Lombardo, 272