The most important thing is to become able to render an acute as a high pitch before the ictus, with the pitch then being or becoming lower on the ictus. Then to distinguish this from a circumflex on the ictus (implemented as a fall). If you can do this, most of the rest is a matter of detail, such as grave as a rise, but less of one than an acute. Falling pitch on the ictus is a general feature of the cadence, but it is a consequence of the way the accent system works, and not a rule per se. People trying to implement the pitch accent often put pitch excursions here and there in the line, without actually getting them to show up in the correct places (my own performances also tend to have various kinds of timing faults). Catathesis ought to happen naturally (however the details of the performance often vary from what the rule itself would seem to dictate, which is not necessarily a problem since emphasis can override catethesis), and anathesis and secondary rise are also easy. As mentioned in the pitches section, the rules say when to go up and when to go down (in many though not quite all cases), but not by how much, so that's up to you. My intuition is that there is a correct choice, but there is at present no known rational basis for this (however close investigation of Ancient Greek musical modes might reveal some).
Then comes the matter of rhythm. At first I just stressed the ictus and let the rhythm take care of itself, not worrying at all about midline caesuras, etc., but with concern for syntactical boundaries (pauses after clauses, etc.) There are quite a number of people around who can do good rhythmical readings of Homer, so this doesn't seem to be a major difficulty. I was well above average in school at reading Greek with naturalistic phrasing so that it sounded like a language, part of how I got that way was a lot of rehearsal of the lines I calculated that it would be my turn to read out loud; I don't know what techniques other good readers used to aquire their skills.
My original method for learning to do Homeric lines with the pitch accents was to first fixate the rhythm of a line with a standard (stress-on-ictus) rhythm (I had previously memorized all the lines), phrasing on a strictly intuitive basis, and then trying to get the pitches 'right'. What was 'right'? At that time (1989-1002) I had not encountered the Devine and Stephens rules, so 'right' was (a) hi on the acute (b) falling on the circumflex (c) hi-ish but not fully hi on the grave (d) a subjective property which I'd describe as 'sounding like something'. (d) is of course the mysterious ingredient whose nature and origin is difficult to explain. It is basically I think a combination of anathesis and catethesis. Catethesis I probably picked up from having recently taught a field-methods course on a language of Uganda called Ateso, which has a tone system as well as what sounded to me like an independent stress system, and so was a sort of living model for reciting the hexameter, while anathesis was a guess, partly inspired by the appearance of the nice effects of `terracing' for a sequence of graves. It's not as surprising as it might seem that a transfer from an African language plus a guess would turn out to get independent corroboration, since the way in which pitches interact with grammatical structure and phrasing in human languages is heavily constrained by language-universal principles ('Noam Chomksy's Universal Grammar'; see Devine and Stephens for details (a lot more than you probably want to know)).
After a few months, most lines started coming more or less automatically, although requiring some rehearsal time, but not needing a rhythm-first, pitch-second approach. I will still stumble on some lines in sight-reading, even after about 12 years of doing this.
However at some point it is useful to learn the standard doctrines about where the caesuras are, since the first phrasing that one comes up with is not necessarily the best, and the rules provide a basis for coming up with specific alternatives. Some explicit thinking about syntax can also help. One useful point is that verbs and their objects (especially preceeding objects) are more tightly bound together than subjects and following verbs, so after a subject is a more likely place for a pause than after an object (Devine and Stephens give some metrical evidence for this). Also there is no reason to think that nouns and following attributive adjectives form any kind of sacred unit, since adjectives can be afterthoughts ('get some green beans - fresh ones'). A random point is that I find that I have a persistent tendency to want to put phrasing boundaries between a major word and a following mén; I don't have a consistent response to this temptation, maybe it's valid.
It is also important to realize that 'stress' as understood in English is not really what is to be placed on the ictus, since English stress is composed of three components, higher pitch, longer duration, and greater loudness (intensity). Since the pitch is being controlled by the accent, the 'stress' must be composed of some combination of greater intensity and length; Devine and Stephens argue that increased length is what in fact characterized the ictus.
A final point concerns the consonants and vowels. Vowels are highly variable in language change and between geographic regions. Since Homer was performed across a wide area for centuries, there well could have been a lot of variation in the vowels (but on the other hand maybe the competitions kept things standardized to some extent). My rendition is for the most part hopefully not too far off from 5th century Athenian, except that the ou probably ought to be a long back upper mid vowel rather than a high one.
Consonants on the other hand are more stable through time. Furthermore I think it's actually important to do the aspirates as aspirates rather than fricatives, and zeta as 'zd'. The reason is that I think there are percussive rhythms, and subtle timing effects which the fricative pronunciations will subvert (if you recite in a whisper, the percussions of stop releases will be more prominent relative to other things, and the effects more noticeable). However if you already use fricatives, you should keep doing that until you've got the pitches under control, because that's much more important, & don't worry about stops versus fricatives until you can 'hear' the necessity for changing.
I suspect that it is necessary to actually memorize a substantial amount to get the hang of this, at least a hundred lines, and much more for even minimum competence at sight-reading (I've memorized between a quarter and a third of the Odyssey at one time or another, although I can't get more than about 1000 lines actually deliverable at any one time). I think it is also completely useless to write out scansions, etc, this is an ear-mouth-and-lung based activity, not a hand-and-eye based one.
As a final remark, I'll note that for susceptible individuals it can be fairly overwhelming if you 'get it', so it might be wise to avoid trying during times when you have urgent real world issues to deal with. And perhaps a word about audience tolerances: about three minutes is probably a good maximum for audiences that don't know any Greek, there will be some who want more but they won't be the majority. For audiences that do know Greek, I'd suggest six minutes, except perhaps if you know a lot of professional performers' or storytellers' tricks to liven things up (I don't). To the 'inner ear' it may seem like you can go on a lot longer, but in an overt performance you're likely to 'loose mindfulness' and fall into various boring and repetitive practices.
Created by: Avery
Maintained by: Avery Andrews
Last modified: 22 July, 2002