means of this system Homer had at his disposal a noun-epithet combination to fit all forms of the names
of all his main characters (as e.g. nominative Hector, accusative Hectora). What is more, he had a
different noun-epithet combination to go in each of the main sections into which the line might be
divided. For example, the last six or seven syllables of the line, often form a section
after a verb (such as the very common prosephe, 'he/she said') which needs to be filled with a noun-epithet
combination. On the other hand, while Homer has a formula ready for all these standard eventualities, he
has by and large one, and only one, for each eventuality. Thus after prosephe Hector is always
korythaiolos Hector, 'of the shining helmet'. The remarkable thoroughness of this system of 'extension'
and 'economy', as Parry called them, must have been the product of an inherited tradition. The
concomitant processes of enlargement and refinement must have taken generations, and must have been
the product of composition by oral improvisation. The development of the diction, passed from master to
apprentice over generations, was practical as well as aesthetic.
Once we have in mind this process of an oral tradition constantly acquiring attractive and useful new
material and casting off outmoded or unpleasing or superfluous material, it may be extended beyond
names and nouns to verbs and phrases and whole lines. For example, in the part of the line before the verb
prosephe there is a whole range of formulae, mostly using participles, which give a tone or attitude to 'he
said'-'in reply', 'standing near', 'greatly troubled', 'looking darkly', etc. It is rather like a system of chemical
elements which can combine in all sorts of different ways to make up different molecules. There has been
a tendency, however-partly due perhaps to the molecular or 'building-block' analogy-to reckon the range
of combinations and possibilities of expression to be far more limited than they are. Parry and his
successors may also have based too much on the analogy of still surviving traditions of oral verse-making,
especially in Croatia; these have comparable formulaic systems, but they are far more limited and crude
than those at Homer's disposal. The possibilities of his diction are amazingly rich, with an abundance of
variation and flexibility and a huge range of subtly differentiated vocabulary and formulaic phrases.
Homer never seems at a loss or stuck in a corner for the right means of expression. His formulaic diction
increases rather than limits his poetic inventiveness.
But while the range and fecundity of the traditional oral language should not be underestimated, neither
should its pervasiveness in the making of Homeric poetry. The inheritance of ready-made elements
extends beyond phrases and lines to whole scenes. This is clearest in the scenes such as serving a meal or
launching a ship, 'typical' scenes as they have come to be known, where whole blocks of lines are
completely or nearly repeated verbatim. But the traditional 'formulaic' scene-shape can often be seen in
sequences where there is little actual verbal repetition. This has been amply demonstrated by Bernard
Fenik for the mass of battle scenes in the Iliad, material which one would expect to be highly traditional.
But the same inherited shaping may be seen in recurrent sequences in the Odyssey. For example, the hero
arrives in a strange place and is at a loss; he is met by a noble stranger who helps him and directs him to
the royal palace. The outline provides a set of directions for the oral poet to work within.
Once more we should not assume that the traditional pattern is restrictive or inflexible. It creates
expectations which as well as being satisfied may on occasion be varied or contradicted. Take, for
example, the slight variation on the usual scene-sequence of the arrival of a stranger at a hospitable house,
which we have when Telemachus arrives at Sparta at the beginning of Odyssey 4. Usually the host